💡┃suggestions-web

[ivory frost]

*Suggestion:

When you generate songs, it would be nice to have a little selection of symbols to choose from to mark that specific generation for purposes.
For example, when you like the chorus, but the rest is off, or the whole song is off, except for maybe the melisma - the songs are stacking up, but deleting would be a waste in many cases.

Instead of editing the title (which can get a little annoying) or upload a cover to tell it apart from the rest - a little option like the emoji picker could be useful, no ?

Clutter is stacking up quite fast - which tempts, at least me, to go tabula rasa and just delete everything, so it doesn't overwhelm.

[half thistle]

you could try setting up separate workspaces and moving tracks into those. may help.

[ivory frost]

I am doing that already, i don't like going over thirty in those.
I've got workspaces for specific projects and then for normal stuff for fun "March 2026/1, /2/..." "April 2026/1, /2..."
but I don't think "liking" or pinning every song where something is done really great is creating more order.
The way the pinned songs are shown in the workspaces are a bit 'meh' - what I want to say, I wish there was a better way to "housekeep" visually.

[indigo rivet]

Getting reports of users uploading their own audio and being falsely flagged for "Uploaded audio contains copyrighted lyrics." Suno needs to give users a way to contest this.

[thorny mural]

Same here.. That was just me humming a melody? There aren't even lyrics lol

[ancient matrix]

i've goten that error on audio only regenerate setions too

[acoustic stratus]

Please improve the Animation, the 10 sec advanced is not enough prompt size for the amount of credits used on it. i just spend 900 credits on it and did trash it all. while my subscription somewhere else allows me to make a certain amount every day for free.
which make me doubt the service that offers here.
on the other hand, it would be cool if we could upload a portrait and it could make lip sync to a entire song with slight movements.

[slow walrus]

Profile is now optimized for 4:3 monitors. xD
Just 1 pinned song on front, only total 4 pins? 🥴

[slow walrus]

suno has the worst broken moderation filters you can ever imagine

theyre filtering in profile page me adding the genre "hardcore"
They censor in comment section totally normal sfw german words. This moderation is absolutely abysmal

like really shockingly bad. When they plan to fix this?

[vast isle]

can we get an option to bulk download songs

[fallow jacinth]

new layout is lame. put the three dots back on the songs page. made the UX Worse. Show more than 1 featured song.

anyone reading this..please, know that.....
using carousels in UI/UX has been proven to be a bad design pattern for almost a decade

Feedback: Please stop using carousels in the UI.
Carousels are proven to be one of the least effective UI patterns for actual engagement. Usability research from the University of Notre Dame found that 84% of clicks stay on the first slide, and only 1% of visitors interact with them at all.

By burying features or information in a carousel, you’re basically making them invisible to the majority of the community. It adds friction, causes "banner blindness," and is a nightmare for accessibility. Please consider a "bento-style" grid or a static hero banner instead so we can actually see what’s new without having to hunt for it.

[slow walrus]

yeah small cover images, design for 4:3 monitors, only 1 pinned song at focus, only 4 songs total
The only good thing is that my profile banner is now showing the persons fully instead of cut off in half, but the rest .... meh ...

[round spear]

Please bring back the magic stick for simple generation (might be double posted - also posted in Android since it was a different platform)

[young zenith]

DON’T mess up our front page… GUYS, what are you doing? Are you all mad? FFS.

You killed Trending, the normal homepage — no more cool playlists, no more genre songs… just this garbage “mood” stuff. What the hell?

Are we some kind of toy for you? People are paying $30+ every month just to get treated like this?

JUST WHY?

I’m not on a phone, I don’t have a 9:10 screen… can someone PLEASE check things before implementing stuff like this?
i am so done ....

[fallow stream]

just wanted to say thanks to the team for adding the credit display toggle. it's reassuring and appreciated.

[glass barn]

Manually picking how long you want this song

[final valley]

... neither good for users nor for creators. One of the worst UI/UX decisions ever on SUNO

[delicate solar]

It is also happening in my end, so UI Change

[heavy marlin]

The new UI/UX on Suno Profile is super ugly. You forced me to speak up about how horrible it looks. Better yet why not ask first before changing it?

[teal shore]

i think one of the best suggestions they can receive right now is, revoke all the changes in the past 2 years they've made to the site itself because when i joined this site it was absolutely booming, and now this site is literally a graveyard (in all 20+ servers i'm in, including this one) with a UI fit for a 640x480 monitor from the 90s, and it's quite worrying. it feels like it's only a matter of time til i wake up one day and there's no more suno because that's what has happened to a lot of video game development teams i have bought games from where they rejected user feedback/critisizm, and now they no longer exist because they did their own thing and ran their business into the ground from silly mistakes, i just hope and pray this isn't what is going to happen to suno 🙏 some steps forward would be a great start

[wintry dew]

Put remaining credits back in the sidebar

[drifting widget]

Guys i can’t create anything and im emailing the support and no response.. is there any help ? I paid an annual subscription and im really stuck ..

[distant sigil]

Suno support takes time to reply back plus they’re backlogged too especially over the weekends. It takes up to three weeks, other option is create your thread inside the ticket help channel follow the instructions there https://discord.com/channels/1069381916492562582/1479483412250366054 try to have patience with Suno support until they reply back.

[distant sigil]

You can toggle the credits on like how it’s shown in the post
#🚀┃updates message

[wispy hemlock]

In version 5.5 of suno , 90% of creations automatically create a "live" audience without specifying it in the prompt.

[slow walrus]

you should add a question mark or some sort of random indicator to this image.

Because I found out, it's random which song is presented first, or better said all 5 are random picked from that carousel

[arctic python]

You now have to go to your account page and toggle "show credits" option to view them in the sidebar again.
Which seems stupidly unnecessary as everyone is going to end up doing this and they should've just left the credits visible as always.

[odd wraith]

Hiding the credits is something that more than a few people were upset about, not sure why it had to be changed. Everyone i know isn't happy with the new home page.

On the homepage your only able to view like 15-20 of the artists songs even when you hit the option to view everything it physically won't load past the first page so its impossible to look through someones catalog of published material.

[winter island]

Today I tested the style generation feature. In the lyrics window, I specified that the verse should be a male voice and the chorus should be a female voice. The result: it didn't work. Many of the generated versions were unusable.
​In the style window, the system doesn't seem to recognize specific details and generates music based on a simplified scheme.
​My Suggestion
​I propose adding a symbol-based system (like arrows or pointing fingers):
​Upward arrow: To pull the note higher.
​Downward arrow: To pull the note lower.
​Straight line: To keep the note flat/steady.
​If it is possible to implement this in the code, it would be a huge improvement. Otherwise, all versions sound monotonous and don't match the creator's vision.
​Technical Issues & Variety
​Recently, I’ve noticed a duplication bug: the song cuts off in the middle and starts generating the same part all over again.
​I would also love more variety in musical and vocal styles. When listening to songs from other creators, I noticed a lot of similarities—the sound is quite repetitive and "cookie-cutter." This is a significant drawback.
​On a Positive Note
​Thank you so much for the opportunity to realize my lifelong dream, even in my later years ❤️🙏 I believe everything will be great as we move forward together! 🙏

[haughty dome]

For Custom mode I don't get why we have to upload files. Why can't you just use songs we already have in our account?

[haughty dome]

You can unhide credits.

[summer lion]

I wish to make an automatic full music video based on uploaded art and the song be true

[halcyon nexus]

🐞 Bug Report: Personal Library Search Returns No Results for Valid Queries
Product: Suno Web
Feature Area: Personal Library → Search
Severity: Medium (impacts usability and trust in library)
📌 Summary
Searching within my personal song library frequently returns “No Songs Found” when using short queries (1–3 characters) or even valid prefixes of existing song titles. This occurs despite songs clearly matching the search input.
🔁 Steps to Reproduce
Go to My Songs / Library
Ensure there are songs with titles starting with a specific letter or word (e.g., “Take Care”, “Alone Again”)
Click into the search bar
Type:
A single letter (e.g., “T” or “A”), or
A short prefix (e.g., “Ta”, “Al”)
Observe the results
❗ Actual Behaviour
The system displays “No Songs Found”
Matching songs that clearly exist in the library are not shown
Results may feel inconsistent or incomplete
✅ Expected Behaviour
Search should return all matching songs from the user’s library, including:
Titles that start with the query (prefix match)
Titles that contain the query
Results should update instantly as the user types (real-time filtering)
Even 1-character inputs should return relevant results
No valid matches should be omitted
🧠 Additional Observations
Issue may be related to:
Incomplete or delayed indexing of user songs
Minimum character thresholds applied to search queries
Backend-only search without client-side fallback filtering
Newly created songs may be more likely to not appear in search
💡 Suggested Fixes
Implement reliable prefix + substring matching for personal libraries
Add client-side filtering of loaded songs for instant feedback
Remove or reduce any minimum character limit
Ensure all user songs are consistently indexed and searchable immediately after creation
Add fallback logic to display closest matches if exact results fail
🎯 Impact
This issue makes it difficult to locate songs quickly

[kindred hull]

Not sure about how Suno manages the "My Taste" feature, but it seems that it tends to generate less varied and less innovative songs. I suggest giving the option to turn this option fully off, or including a slider as a the Suno parameter, form "Match 100% to My Taste" to "Match 0% to My Taste".

The Weirdness and Style Influence sliders used to be useful for generating more varied output, but lately they don't seem to effect much on the style of the song: the same prompt with big Weirdness and low Style Influence still tend to generate very similar songs.

A feature to convert old songs to the prompts effective for the newer versions would also be welcomed. Prompting in newer versions is very different from prompting older versions. You can download and upload a song to see a prompt fit for the last version, but this procedure is quite slow.

[unreal spoke]

I am having this exact issue, My styles was already turned off but my taste was still populated. I have deleted it all but apparently needs something in there to save it. bad news is it hasn't changed a damn thing. even when I prompt for a completely different genre but with the same lyrics it still keeps the vocal phrasing and any instruments from the initial prompt. Not using cover or remix, cleared browser cache and still it will give you the same song with minimal changes.

[kindred hull]

Have you tried entering something in the My Taste text box (something like "innovative" or the like? I did it on Friday, and it seems that improved at least a bit.

[unreal spoke]

Didn't think to try that, I just put do not use. I will try something like your suggestion when I try again. Ty

[kindred hull]

Please let me know if this works for you! This issue has been making me mad. I used to like ALL the songs that Suno generates, but now I like NONE of them!

[knotty rain]

I've been putting off posting this very issue, as I deal with the (speaking for myself) more annoying matter of Voice management.

That said, I've also been putting off making more Custom Models because of this specific issue.

[rare rain]

Suggestion: Option to add ISRC or other fingerprint for a song when creating (perhaps in "More Options"). This allows an artist to cover or remix their own published song using Suno.
Alternatively, add ISRC/fingerprint values in your own profile (hidden to other users) to function as an "exception list" while doing copyright checks.

[woven burrow]

@anyoneherethat IS A MOD, STAFF, ADMIN, OR THE OWNER.

Please amend your server rules (either add, make a change. I'd prefer just establishing a new rule). Where users are mandated in ANY/ALL channels (especially general and suno school) that if a user is going to post a link of a suno song of theirs; they must do the following:

1. BRIEFLY and concisely (option to go in-depth but not a requirement) state the genre, and your objective (i.e., do you want constructive criticism, ACTUAL suggestions that will be incorporated - meaning the mutual exchanges aren't just people talking to a wall, etc.) when you post a suno ai song link in ANY channel

The entire purpose of this rule is to maintain austerity, test the user's critical thinking skills, and not just be "another Soundcloud". I'll give an example of this Amendment below to demonstrate how the interactions should occur:

DO NOT DO
" Username456: "Hello everyone, [suno ai music link], please listen and rate"

WHY?

• this is "Soundcloud-isms/Soundcloudisms". This is no different than you going to a Costco, and someone is trying to sell you shutters and all you hear is "Hi! Our shutters are 5-star, please look and rate and consider buying"! You're just being a Soundclouder, not someone who actually wants to revolutionize the music industry/NOT saturate it

---

What should happen at ALL TIMES if this is the case
" Username456: "Hello everyone, I'm making an industrial baroque pop rock song, and I would want someone who enjoys this genre to listen to it. Here is the issue [xyz argument/where Username456 needs help/feedback], see link below!:

[suno ai music link]"

======
My goal is to problem-solve, create a MORE genuine community and not a repository or Soundcloud 900.0, not beat users with a cudgel with red-tape.

If my suggestion falls upon deaf ears, expect more people just larping as monkeys going "ook ook! Here music, listen, rate". I don't believe my suggestion is unreasonable. It's robust, will teach users to think critically about their questions to ensure accuracy, and overall create a positive Suno ai experience

[wintry dew]

Thanks for responding. I worked out how to do this, almost immediately after I posted. It seems like a bit of a sneaky one so that you get through your credits without realising, and SUno is hoping you'll upgrade your account to finish what you're working on when the credits run out.

[wide compass]

Hello, please bring back the credits remaining count in the top left corner when making tracks. Not sure why anyone thought it was a good idea to remove that. Just realized you have to enable it in the Account settings tab in case anyone had simliar issue

[west elk]

And what do you propose is the method of enforcement for users that don't do all that? Making up rules is easy. That doesn't make you a "problem-solver." Also, just declaring some sort of punishment for those that don't follow the rules isn't going to help (at best) and is going to drive people away (at worst). So, whatever enforcement mechanism is implemented would need to be more carrot than stick.

Also, fwiw, this really sounds a lot like a "you" problem, than any sort of systemic problem with Suno or the Discord. Yes, those things would be nice, but I think you've presented is a rather more extreme situation that it really is.

[west elk]

Also, who the hell do you think you are? Talking about proposing "Amendments" and "testing user's critical thinking." The self-important audacity.

You're brand new. And we know you're new. Because all of the "problems" you identified have been around for a long time. If you'd been around awhile, you'd know that. And none of the catastrophic things you foresee have happened. If you'd been around awhile, you'd know that, too. So you're brand new. And you have the balls to come in here like you're gonna "fix" the place.

Worse, you can't even articulate your point yourself. That's pure AI-generated garbage you've posted. And we know it.

Please tell me how "critical thinking skills" are actually "tested" by anything you proposed. Then tell me why "critical thinking" is important in the context you are speaking about. You can't, because it's not.

Also, what TF are you proposing an "Amendment" to? You don't seem to understand half the words that appear in your AI-generated puffery. Next time you want to make a suggestion, write it yourself, or proof read what Chat GPT hands you.

And you come in here in all caps talking about @ TO WHOME IT MAY CONCERN, ALL ADMINS, MODS, AND DECISION MAKERS, like you're some sort of divine oracle from on high. Get over yourself.

[west elk]

Yeah, it's a terrible idea. Good news though, there is an option to re-enable it, if you have any of the premium accounts.

[west elk]

Making up problems to solve isn't "problem-solving." You want a "MORE genuine community." What makes you think it isn't genuine in the first place? You've made up a strawman problem so you can act like the newbie hero that just swanned in one day and "fixed" everything. Oh my gosh, you must be so brilliant.

Your "suggestion" is neither robust, nor will it "teach users to think critically about their questions to ensure accuracy," nor will it "overall create a positive Suno ai experience."

You clearly don't know what "robust" actually means.

And you clearly don't know how to proofread. You've literally pasted nonsense spouted by Chat GPT. "critical thinking?!" Really? About checks notes "asking questions." What questions!? You started out talking about posting information about songs, not asking questions.

And right now, the two biggest detractors to a "postivie Suno ai experience" are the Suno team making terrible design decisions with respect to the UI, and people posting garbage like you have. Neither of which will be solved by your proposal.

Next time you post AI-generated sh1t, proof read it to ensure it at least makes sense. Good grief.

[tropic dust]

man spoke nothing but facts

[woven burrow]

Found the Soundcloud user

[woven burrow]

Please read the argument below:

"If my suggestion falls upon deaf ears, expect more people just larping as monkeys going "ook-ook. Here music, listen, rate"." pretty much 34/7/365. You just re-created Soundcloud.. just saying.. before you make any more -9,999,999,999,999^999 IQ arguments.

If you unironically think that's going to create an engaging, critically-thinking, and helpful community; consider contacting the SSA, and living off of SSI due to Incompetency. This goes for @Esoterick (who ironically has a name, meaning "intended for or likely to be understood by only a small number of people with a specialized knowledge or interest."... and unironically has an IQ lower than a butterfly)

And @Esoterick is supporting the status quo of people.. who have little knowledge, and a mass amount of people doing the same thing without questioning it.

Again to both @Patrick and especially @Esoterick consider contacting the SSA, and tell them that you have a sub-70 IQ irl, and you substanitate that by saying "I support netizens going on Suno ai discord and just posting their links, saying only listen and rate". You'd already skip all medical diagnoses and psychological evals, and IQ tests, blindness tests, and reaction time tests to live off SSI because you're beyond Incompetent

[subtle ocean]

How can I clean the trash with one click, one by one, to get rid of all this garbage?

[woven burrow]

When you navigate yourself to a local govt. testing facility, so you can pass your GED you flunked since age 40

[subtle ocean]

why are you so cheeky? there's a lot of trash in the library's trash can, and to permanently delete it from your account, you have to click on it one by one or mark it one by one.

[woven burrow]

Because my job all day is to do the thinking of fully grown adults. When no one notices issues but only I do, it's a sign leadership needs to enter and provide coaching and mentoring.

when genuine morons (like those esoterik and patrick clowns) think this Discord should just be another Soundcloud (but they would NEVER use Soundcloud themselves), is a sign there is NO leadership, NO critical thinking, and it's time I'm elected to provide solutions

[subtle ocean]

What does Patrick have to do with it? I asked a completely different question )

[woven burrow]

I eradicate imbecility, I erect leadership, I solve problems, and I proliferate correct behaviors.

If you're against this, consider living in San Francisco, where you can meet like-minded individuals

[subtle ocean]

I asked a question about the library and the basket.. please delete him from the chat

[woven burrow]

You can drink soy, worship homeless fentanyl addicts, dye your hair blue, get 5 quintillion piercings and tattoos, and no one will say you're doing anything wrong (besides your conscience)

[woven burrow]

Let's revisit. You typed "How can I clean the trash with one click, one by one, to get rid of all this garbage?".

no context.. no nothing.. just immediately after my reply. In case you flunked the 1st grade English and have no idea how Clauses work. You asked a question, indicating "all this garbage". I assume you're talking about everything you don't like which includes my arguments.

(not that mass censorship like your question, has ever worked anyway)

Did I misunderstand something about "all of this garabage"?

[subtle ocean]

It's not just your problem that's being discussed here.. I asked about the library in my account and how to clean the trash out of it in one click))

https://tenor.com/view/sealyx-naked-gun-the-naked-gun-facepalm-facepalm-meme-gif-11743768702765397589

[woven burrow]

This is #💡┃suggestions-web .... not an FAQ.. As far as I'm aware you have to just click the thumbs down on every song you don't like

Suno doesn't have a mass sort function yet

https://tenor.com/view/джона-хилл-jonah-hill-facepalm-рука-лицо-долоня-обличчя-gif-7167197686338732060

[subtle ocean]

there's just a lot of garbage, and that's the problem. so I asked

[woven burrow]

You just have to manually click the thumbs down so far. I'm guessing the programmers at Suno presupposed if a user "disliked" or didn't like a Generation, they wouldn't listen to it again and it's removed from your account afaik

[subtle ocean]

suggestion - make the clear all button

[woven burrow]

Clear all? Even songs a user "likes" or doesn't intend to wipe?

[subtle ocean]

It's removed from the library, it goes to the trash.. and there's everything I've deleted over the course of a year.. it's a lot ) it's Suno's memory, it's expensive, and I don't need it! )

[woven burrow]

I see, you want function to mass "Move to Trash" and have it execute a perma-delete

[subtle ocean]

so that when you go to the trash, you can permanently delete everything with one click.. that would be the correct way to say it

[woven burrow]

right right, I see now

[subtle ocean]

thank you for understanding me correctly)

[woven burrow]

the Trash is a consolidation of info, while the execution of trashing all, is needed in incorporation

[clever harness]

ctrl+a (select all currently visible) + ctrl+backspace (delete) = bulk delete max of 250 at a time

[subtle ocean]

Thanks, I tried Ctrl+A and the entire text was selected.. I'll try again

[molten bluff]

The reason that the thumbs downed songs disappear from the library list is because of filters, which can be disabled above the list. Thumbs down just hides them in the default filter settings.

The move to trash feature is very slow, especially when having to move many songs or sounds from studio projects into the trash one by one.

[bleak lynx]

Step back a moment. The 'Library' contains various artifacts and the information is primarily a tabular display (on the web at least)
1 For tabular displays add a chooseable check-box (all, likes, dislikes,etc) as first column in table header which in affect can help selective bulk deletes
2 Add a Delete all selected in the table header that is only available when 1 or more table entries are selected (see 1 above)
3 Make all Library tabs (i.e. Songs, Playlists, etc..) tabular displays (or option to show in table) and add 1 and 2 above
4 If a filter changes, un-select anything selected
5 In the actual Trash view, 1 and for 2 change Delete All to Permanetly Delete All

[shy dove]

​Hi Suno Team!
​I’m a young film composer, and I’ve been using Suno to orchestrate my own melodies. The sound quality is professional, but as a creator, I want to go beyond simple song generation. I want to use Suno as a professional tool to build my own music, note by note.
​My Vision:
Instead of just generating full tracks, I’d love to see Suno become a Digital Audio Workstation (DAW) or a VST Plugin where:
​AI-Powered Virtual Instruments: We could "prompt" a specific instrument (e.g., "Cinematic Erhu mixed with an Orchestral Cello") and Suno would generate a high-quality Virtual Instrument that we can play ourselves via MIDI.
​Total Control: We keep the surgical precision of a traditional DAW (placing every note, controlling the rhythm) but with the "Hollywood" sound quality that Suno provides.
​Democratizing Film Scoring: This would allow young composers who can't afford expensive orchestral libraries (VSTs) to have access to top-tier sounds for free or at a low cost, focusing only on their creativity and melody.
​The Goal:
Turn Suno into a workspace where the AI is the "Instrument" and the human is the "Architect."
​We want the Suno sound quality, but we want to play the notes ourselves!
​Thank you for everything you’re doing to change the music industry!

[spring herald]

I have to agree. It is getting to much now. Endless 7:59 songs created. About 3/4 into the song, it will sound if the song is ending. But will restart again with the 1st verse. Or just a 7:59 generated song that does not make any sense. A big bug this is in V5.5. Please fix.

[little sleet]

suggestion - this language filter is insane and arguably microaggressive due to an unsurprising lack of cultural nuance.

[knotty rain]

Funny thing... I rarely get anything longer than 5 minutes, and when I do it's usually because my lyrics are long enough to warrant a length in that neighborhood. I just paged through my works, and the last track to do that 7:59 thing was back in February, under 5.0.

[slow walrus]

I wonder how many reminders and feedback it will take until they revert the terrible new profile page

I wonder which mad person thought like: YES this is better than before!

[arctic python]

Yes I wish they would address this issue, I have had to go back to V5 because literally every gen in 5.5 is 7.59 and not functioning correctly

[slow walrus]

that 8min problem I did actually never ever have. Strange.

But therefore there is some noise in the treble which is quite unpleasent about 5.5. This reminds of v4

4.5+ and 5 are clean though

[south robin]

Bulk-Downloads for my songs

[austere zenith]

can we add a repeat option in web playlists like in mobile? I want to hear my playlist not some random ahh song

[pallid cloak]

The only solution I have found is keeping 50/50 on the weirdness and style. I seem to have found a balance on everything else prompting wise and haven’t had a 7min render / hallucination since yesterday afternoon. 🙏🏻

[potent vale]

How is it possible that the credit is taken from you, the melody is pretended to be created then disappears, the melody and the credit disappear into the mist?

[broken wagon]

Can we have the old profile page look back, at least as an option maybe? When viewing peoples page I can not even see the style of the song without clicking into it first. Nothing about the new look is better in any way.

[pallid cloak]

I’ve been expressing this exact idea with the developers and my own personal community of music production artist for months. Essentially, I think it’s inevitable that this will happen however, I’m glad to see other people expressing the same desire.

[modern summit]

please suno team for god's sake can you please remove the auto style gen while uploading, or at least allow us to edit it? most of the time it's useless and has point with the actual content. also it's so unnecessarily detailed, with elements and terms i don't even understand. can't it be simple? i don't care about all those fancy stuff, it just makes my mind freeze and makes me completely anxious

an example: "J-Pop and J-Rock fusion, Distorted electric guitar plays a syncopated rhythmic pattern alongside a driving drum kit with a prominent snare and crash cymbals, A high-pitched female vocal performs a melodic line in Japanese, The arrangement features a thick synth pad and a melodic bass guitar line that follows the chord progression, The tempo is 175 BPM in the key of G Major, The production is polished with bright equalization on the vocals and a wide stereo image for the guitars"

what is syncopated rhythm? what is driving drums? why all this, who cares? not mentioning, again, that 95% of this is just completely wrong and i can't change it. style summary seems to be completely useless

i'm sorry if sound just like venting or raging, but i'm actually trying my really best to not panic compleetly right now...

i'm not english. i use the website in elgnish cause i don't mind but still i'm not fluent english especially for all that music production stuff. i'm just a casual doing casual stuff, i'm perfectly fine with simple and one word stuff. electro, j-pop, piano, synths, female vocals. i'm ok with that, i don't need anything else so ellaborated, i feel like it just sticks me into one precise type of generation and nothing else. i just want to generate random and different things, not being forced to one rhythm, one voice, one drumkit, one bass, etc etc

[half thistle]

why not just install a DAW and some of the AI VST's out there on the market and learn how to make that cinematic sound, maybe not change suno to something thats already out there in many many forms. just saying.

[shy dove]

Yeah but the more the better plus we wouldn't lose anything.

[half thistle]

Until the make v5 say words properly, or v5.5 not crackle, or stop songs being 7.59, i think let them concentrate on what they've got first, and negate all the other comments above.

[shy dove]

Okay so wen they finish will see.

[half thistle]

sure, but as a company surely it would be a waste of money and resources to make something that's already out there. its too big a risk, but you never know.

[spiral bronze]

Honestly if they just make Studio work, then hopefully they simply allow you to be like "Alright make it quiet for two minutes, bring the energy up, add a lull in the action here, heroic reprisal" and that's all the directing they need to start with. Give us the ability to save some central motifs and melodies and build compositions around them and we have film scoring solved. Give us looping tracks and we have video game scoring done. Let it jam alongside a human with a MIDI controller and we have everybody else covered. Currently if Studio can't even work with stems properly then the rest of it has to stay longterm goals though.

[slow walrus]

I want to keep it. It makes it easier to understand suno which phrases it understands
But an editing possibility I fully support

[winter blade]

Hi Suno team. A suggestion to maintain voice consistency would be to fix a window to select the singer's voice above the audio options, forcing the platform to use the selected voice instead of a new one. Thanks for all.

[tepid laurel]

Man.. I just tried gens at various Audio Influence and Prompt Influences strengths

And they literally ALL produce the same exact gens. There is NO DIFFERENCE.

These sliders don't do ANYTHING.

i can still hear the cover audio Exactly. There is no change to the audio being covered.

You would think with audio influence at 0.1, or 0, that you wouldnt even hear the original track anymore.

but no. Its the same exact track in the output as in the input.

Why

[spiral bronze]

When using remix, setting the influence to 0 still pulls the song's structure but doesn't sample the audio. Style Influence acts weird depending on how different your new prompt is from the prompt for the track you're covering, but it seems like pushing style influence up while audience influence down makes the source track weaker in the final output ONLY IF weirdness is set high enough. If weirdness is down low, then it doesn't change the structure of the audio much. High audio influence only seems to act like a sampler in Remix.
Now in "Inspiration," setting the audio influence to 0 is effectively like setting it to 5%. It doesn't adhere to structure as much, it seems to try to grab instrumentation and phrases (musical, not words) from the source track(s). With weirdness down it still changes things, but it's much more what I would consider a "Cover" because it still is recognizable in the end.

The only reliable way I've found to actually change tracks using the remix or inspiration features is to either generate multiple of the same song and mash them together in inspiration, or to drastically change the genre of the song in remix and then remix it again.

[tepid laurel]

weirdness was high, .9. it still sounded exactly like input

in inspiration it would actually be heavily altered but also sped up because it cuts out everything that doesnt have a word in it

[spiral bronze]

I suppose it also depends on genre and what exactly you're asking for, what the input is, etc.

With what I do, weirdness above 0.6 usually completely breaks things and it starts sounding like I'm randomly scrolling the tuner on an old radio.

[tepid laurel]

well like.. it shouldnt know about the voice, synths and beat of the input if audio influence is 0. it should be as if i was generating something from scratch. but it isnt

[spiral bronze]

Oh are you using voices at the same time as remixing?

[tepid laurel]

When setting audio influence to 0 it should basically be turning up noise and turning down the input audio

yeah

but the voice isnt working

[spiral bronze]

Okay that's what's wrong. That's bugged right now.

[tepid laurel]

its still the cover audio and not the voice

[spiral bronze]

Remix + voices doesn't work correctly since they both use the same slider.

[tepid laurel]

I was thinking that can't possibly be true

like maybe voice just has one single setting

and turning down audio influence would mean that the voice gets to be stronger

[spiral bronze]

Ideally you'd be able to put the voice influence up high and your cover track influence down low, but right now since they're tied to the same setting you have to work around it.

[tepid laurel]

but maybe not

Yeah

[spiral bronze]

Turning influence down makes the voice weaker.

[tepid laurel]

I agree

ah

well that makes no sense..

[spiral bronze]

What I've been doing to get around it is either making a track with just the voice and the same lyrics and then mashing it up with the cover song, or you can sort of get around it by cranking both influence styles up to max and turning weirdness down to 0.

[royal venture]

I would like the ability to make a track, and make a new track that is a direct continuation of the previous track i already made, seamlessly, or with seamless transitions. Why? I want to make orchestral scores that are long, like 10 mins long, for VGM's. I want to make full tracks that evolve and transition over time. Currently our only options are covers, which has too small a context window, or making a full song in a batch of many really small song pieces, which dont include the full context of the entire full song, but only the context of a really small piece. This means I cant for example bring a motif back from an earlier section into a future section, without just stemming, cropping, and pasting it a bunch, and I likely wont even get it to match up or align well with the audio I have later on in the full song. It would be a lot easier If we could context an entire track, or create full songs, and extend into a direct seamless continuation of that song, with marked reference points. The extend function sucks, is not good for this, and is not helpful at all in this scenario either, the context window being one example of why it sucks, not counting other issues with the extend function.

[frigid bough]

The new button: "Personalize style prompt to match your taste" it's an absolute piece of shi... Please put the button: "Improve your prompt" back again...
The "Personalize style prompt to match your taste" It's a big shi... because if you want create different styles of music, this cra...p feature will mix the styles in the same prompt creating authentic vomit noise...

[tepid trench]

Click on your profile picture on the site and select My Taste, then turn off My Styles (the option shown in this screenshot) to make the magic wand behave like it used to.

[wooden lance]

i would like to have a collab option between people, so just like on spotify a song will show up for whoever was involved

[iron wedge]

Why users have to click 2 times just to download a WAV? First click 3 dots menu, then Suno prepares it and there's window with another Download File button (2nd click). I know it takes time to prepare the file but wouldn't it be more simple just spit a notification that it's preparing the files and go straight to the ask the hardisk location windows ? Anyone aggre, please like this question !

[delicate yarrow]

The new profile page is a significant downgrade from the old one. Dishearteningly so.

NEGATIVE
-It hides all the pinned songs in a row so that someone would have to browse them separately. You are no longer displaying the songs you are proud of front and center.
-Pinned songs, you used to be able to see the animations. You could browse the songs and animate cover art. Now cannot.
-It still hides all of the likes/comments. Even more so now that pinned songs are hidden in a list.
-It shows voices as big icons, but songs and playlists as small ones. As if voices are the most important thing on the page.
-It splits the about from the header for no reason, hiding social links and genre tags at the bottom of the page?

POSITIVE
-The only positive change is that it shows the number of songs at the top now.

[tepid laurel]

The Custom Model is too weak and doesn't work.
Personas are too weak, and do not work. Neither Vox Nor Legacy.

And now personas don't even have any effect on inspos.
Before inspo was the last remaining thing that voice/persona still had a strong effect on after Custom models have been watered down into oblivion.
Now even if you turn audio influence all the way i dont hear anything from the persona in the inspo.

• The Custom Models are too weak.

• Personas are too weak. They do not alter the voices you are covering, they do not alter the instrumental, they literally do nothing. They are empty.

• **They cause no change to the input audio in cover mode or inspo mode. **

• Neither of these are having ANY ability to alter the input audio in both cover mode OR inspo mode.

• **And in Sample mode you can't even use personas or voices. **

[warm gazelle]

• I am OK with the fact Songs includes also featured ones, but yeah, hiding like and comment numbers seems so unnecessary. Why?

[tepid laurel]

ok it seems to work again

the persona that is..

but only when using the persona on a new track from scratch with nothing else..

[late elm]

I prolly have... thousands of songs now with the 7:59 song time. I am positive maybe 25 actually were prompted to go that length. The rest? .....stolen credits IMO. You guys should refund credits for 7:59 length'd tracks.

[late elm]

You should DM Superrondotcom on suno or superron21 here on dc fam. He has 2hour long songs. He has a 10 song album that is 8.5 hours. All suno genned. It can be done.

[untold grove]

the new style design makes everything too small.

[drowsy token]

Hi Suno team,

I'd like to report a recurring audio glitch I've been experiencing with the Cover feature on V5.5 over the past few days.

Issue Summary:
When generating a cover using V5.5, the output frequently contains sudden, sharp vocal cuts or interruptions — like an abrupt "pop" or "click" in the vocal track. This happens at seemingly random points during the generated audio.

Details:

• Feature affected: Cover generation
• Model version: V5.5
• Frequency: High — occurs in the majority of my generations
• When it started: A few days ago (this was not happening before)
• Prior behavior: The same cover workflow worked fine on V5.5 until recently, with no such glitches

What the glitch sounds like:
The vocal suddenly cuts out or "pops" mid-phrase, almost as if a clip was spliced incorrectly or a processing artifact was introduced into the audio pipeline.

What I've already tried:

• Regenerating multiple times with the same settings → glitch still appears in most outputs
• Using clean, high-quality source audio for the cover
• Adding style prompts like "clean vocals, smooth" → no significant improvement

My assumption:
This may have been introduced by a recent server-side update or model adjustment on your end, as I did not change anything in my workflow. It seems to be a regression specific to the Cover feature in V5.5.

I would greatly appreciate it if this could be investigated. Please let me know if you need any additional information, generation IDs, or audio samples to help reproduce the issue.

Thank you for your time!

[royal venture]

Probably a feature to combat bots mass downloading wavs, or websites outside of suno fishing download links.

[earnest jetty]

Carousel thing for pinned songs isn't good.

• shows just 1 song, but barely saves space compared to what was able to fit 5 before?
• adds user action to scroll through (noone's gonna do that - good UI = effortless UI, less clicks to make = better)
• it just does not look good, what is all that wasted space about?

Carousel for the sake of carousel is never good idea, it has no purpose there.
At least let people pick which way they want to display their pinned songs if you really want to support several layouts (doubt)

This is bad too (picture).

It should not take more space if nothing is present.
Look at youtube as example. No background thing still looks fine - no wasted space for 0 reason.

• why additional information at the bottom when header is basically empty now?
  ...so many questions about design choices there...

[sage hinge]

Is suno working properly now?

[covert hawk]

Same Error code 504 my side, South Africa

[random sand]

Sam here Australia too

[tardy sierra]

They're working on the fix as we speak

[midnight yarrow]

when will they release the servers

[versed forge]

Why doesn't host working? Is anyone here who know?

[tardy sierra]

The office Hamster escaped XD

[random sand]

I wish they'd stop changing it so much, version 4.5 was so much better than the last 2 upgrades, anyone else notice your left over credits disappear now

[quiet aspen]

same in france

[golden moth]

Allow us to share our music threw other ways not just sunos direct

Some of us put in more effort outside suno and dont have the final product in suno

We should b s allowed to use drive or link it from where its released to

[rare grove]

I do use a mobile phone, but I also got this absolutely awful new design. This probably means I won't be publishing anything anymore and my subscription is expiring.😐😡

[minor pine]

You can still use 4.5...

[bright gate]

the displayed leftover credits can be activated and you got them back #🚀┃updates message

[tardy sierra]

I think most of my favourite songs were made on 4.5

[random sand]

The page layout was better too

[tribal pewter]

I would like to call upon everyone who dislikes the new profile layout on desktop to express our dissatisfaction about it in this channel and to call for a revert to the old profile layout on PCs: #🚧┃suggestions-layout message

[tribal pewter]

I completely agree. Let's take joint action to notify the team responsible for the layout that we want the changes rolled back.

[tardy sierra]

I agree, it's worth making our feedback know, just as long as people stay respectful and feedback doesn't turn into angry mob abuse.

[lapis plank]

Good morning.
I know there's currently some issues but, I figured this creates a great time for me to ask (hopefully 'here' in the right place) for a feature(s) - being the web suggestions channel / tab.

Some of us (I can't be the only one) have built a 'brand' using the tools you offer through Suno - vocals, instrumentals, music...
...and I wonder if there could possibly be some features put into place that protect that / those brands?

...things like:

• Publishing songs but not necessarily to the public. (I know this contradicts how most people use Suno but, for some of us, it'd make a huge difference).
• Making sure that the changes you make to Suno itself don't change the personas / voices we've created and save...ie: tone, grit or being attached to songs that I now have to regenerate due to personas / voices no longer being tagged / attached to them.
• Teaching Suno to prepare and predict habits for users is a great basic feature BUT, people like me, who have created a cast of personas / voices for their brand(s) may find it isn't such a great feature because it continues to incorporate those locked-in habits to all personas / voices rather than just the one it's supposed to.
• When we use a personas / voice for an 'inspired by' song (rather than a cover, when remaking / refreshing a song of our own creation), I choose a persona / voice and click on 'voice' to be used...why does the instrumental that the voice was created from, come through into the song, as well? We didn't click on style / instrumental...that's an entirely different option we could've chosen but didn't click for the song we're using in INSPO...and it's not helpful until it's been fixed properly, please.

...these're the things I can think of right now (coffee's not helping my headache and it's 7½am right now (maybe others who've been building brands have more ideas, as well).

Thank you for taking the time to read this (assuming you took the time to read this) ❤️

[minor pine]

Things start getting sucky when you don't focus on one thing. There are other services for marketing etc.

[rare grove]

I posted several times there, no reaction at all. So I'm reacting...

[lapis plank]

Yeah, there are BUT, the way Suno features work and are steadily growing AND the amount of traction and recognition Suno's gained for those features - I know I'm not the only one with these questions...I'm just the one currently asking. 😉
So, if anyone who has control over those issues / potential changes were to read what I said and possibly have personal feedback directly to me, they're free to contact me directly and offer input but, I don't think suggesting that I use another service just because the way I use Suno differs from how you use it or it's features is a suggestion that works for me, the millions of others who could benefit from these suggestions or the track Suno's travelling towards already. 🙂

...of course, I don't mean to be angry or disrespectful towards you for seemingly disagreeing with me but...your suggesting that I use another service when I'm genuinely trying to help Suno as a whole as well as my use of it, to grow...seems misplaced and brush-offish rather than actual input. ❤️

[tribal pewter]

Exactly, that is why I try to raise the issue even more often and seek allies like you, because otherwise there will definitely be no change back to the old situation.

[lapis plank]

But also, that's my point, though - there're SO many directions people can go with the tools and service Suno offers. Offering many lanes but making features only handle one lane at a time seems...well, backwards IMHO.

[minor pine]

I don't want Suno to go downhill, so I would prefer they don't mission creep. And I hate to break it to you, but no one cares about your brand. There isn't a audience for AI music yet, at least not a real one. as far as I can tell you have 2 groups, the largest are other AI artists liking each others work and are propping each other up in a circle je... well , snake eatting its own tail.. The other group are people listening to AI without knowing they are.

[lapis plank]

Ok, so I read right...no actual input and basic trolling.

I don't need schooled on your opinions so, I'll stop posting here, as I've said what I came to say.

But thank you for publicly showing that you don't understand my post and just wanted to show your distaste for people building a brand - I didn't come to SHARE my brand, to clarify...so, maybe read what I typed again so you actually understand it OR...hear me out...rather than tryna show distaste for people you don't know, in a suggestions channel for a program we all use...you can keep that little tidbit of your personality to yourself, because it helps no one...least of all Suno.

K, have a good day everyone! ❤️

[minor pine]

I didn't mean to be hostile, I thought I was helping you from wasting your time. Can't waste time in this economy!

[tardy sierra]

It would be cool if they made a music video creation tool too all in one big creativity suite space.

[gray barn]

Suggestion: Pro-unlock model

I’d like to suggest a “pro-unlock” ownership model.

Right now, songs created on the free plan feel disconnected from the paid experience. A possible improvement would be:

• Free-plan songs remain tied to the user
• When upgrading to Pro, those songs are fully unlocked (usage rights + ownership clarity)

This would make the free tier feel like real creative investment and make the upgrade feel like a continuation rather than a reset.

It could also be combined with clear Free / Pro labeling on tracks for better transparency.

[soft mortar]

Well... I think Suno regarding "attention to user" have to improve in all ways...
Getting in mind most of us are paying for it, some are paying Studio version (Premier)... Suno should be more friendly with community. Specially regarding credits.

1. In some bugs you lose credits and don't get refund.
2. If a track get stuck, it says no credits are used, but yes, they charge the credits from beginning, they are used.
3. In the case of shutdowns, etc... There is no kind of reward or similar to users (at least the ones who pay)... Something I never seen it anywhere honestly, specially if they affect during days like what happened with v5.5 not long ago.
4. Listen and Reward, a minute you have 10-20 to review, you refresh page and get 0 for days... (The only good thing to recover the credits "wasted" by bugs, and well... YOU NERF IT!)
5. Charging credits for stuff you get "barely" usable (Stems, ejem...) where not got an improvement on bleeding since where? Even that, the track with stems do not sound like the normal one...

There are a lot of things to work on, community contribute on all, but there is no return in that aspect.
Details are the key to keep customers, probably you don't need it right now, but keeping customers safe, make customers more willing to contribute on it. So get it in mind.

[solid cargo]

Hay web DEV team .. stop making the Web Suno interface like your Mobile Phone interface, looks and function like shite ... Now not being able to put your categories in the order you want. No one cares about Hooks, stop promoting them. Minimizing all the work people put into their song cover art and spending their tokens to animate them only to make them small icons, phhhh ... thanks!

[glass barn]

Button to stop the generation before it finishes so I don’t waste credits

[tribal pewter]

I completely agree. I also put a lot of time and effort into my cover images and cover videos, and now that they have become tiny icons that you can't even see, I am really deeply disappointed, so hopefully this decision will be reversed as soon as possible. #🚧┃suggestions-layout message

[solid cargo]

Also, taking away functionality to be able to rearrange the categories to what YOU want to promote. I get having to unifying all interfaces but taking away functions and options to promote what THEY think people want is bad. I would move Hooks to the bottom and make them small icons. Give back the user choice what categories to promote. And if you want to make a cool animations and such , so be it, but give back the user their custrimations. Not everyone using the mobile app to use Suno. The mobile app is bad by its self.

[tribal pewter]

Good point, I also indeed miss the feature to sort the categories myself; personally, I would put hooks at the bottom because I find them much less important. I also find the mobile Suno app really terrible to use, so I mainly use it to check my profile and notifications.

[deep rain]

Now than Suno generates full songs at once without so many errors, I'm having a hard time using 10K credits a month and they don't roll over. 2,000 credits isn't really enough for me, so perhaps Suno can consider a 5,000 credit tier that includes Studio and costs perhaps $16/month and not $24. As is, since I have a life and I don't need to generate a ton to get good takes, I'm losing credits every month because they don't get used.

[minor pine]

I use the credits to experiment more.

[deep rain]

Not enough time in a day for me. I also do AI video and editing, as well as compose music traditionally.

[minor pine]

I think the credits are to keep OCD types with no life from using all the resources. Doesn't sound like you are losing anything and are a tad greedy. Be thankful.

[deep rain]

You're joking, I'm sure. I shuold know better than to type anything on social media and not expect that sort of answer.

[vital abyss]

the new UI terrible compared to the old one

[spiral bronze]

I wouldn't mind a few fewer credits (8,000 instead of 12,000) if they would roll over to the next month or two instead of all instantly expiring on the next billing cycle. I don't like to have to run myself dry just to feel like I'm not wasting credits. Granted, even when you use Suno you're wasting credits because they can't figure out sound quality, song length, how to understand I don't want a loud track, how to design a single website page. It's probably too much to ask for basic quality of life features, better just ask for a new way to make the service have nothing to do with generating music, like making short-form looping videos for pre-teens.

[minor pine]

so you're complaining about their generosity... Someone make it make sense!

[unreal spoke]

Having a certain amount of rollover would be more advantageous overall. Makes perfect sense to me

[spiral bronze]

No. I'm complaining about Suno not being worth putting 12,000 credits into every month just to not waste something I already paid for that I'm not getting because I have to generate the same tracks 50 times to not get screaming vocals.

It'd be nice to be able to ragequit Suno and come back the next month with my credits still there.

[uneven ravine]

Is Suno working on creating individual stems first, and then combining them during the creation of a track? Stem separation quality is tough given how the whole track is created at once.

[teal shore]

what would be good is, if after you split a section in the editor, you can choose to increase/decrease only that section, it would be lovely

[broken mason]

please make other payment method to subsciption like alipay or qr code or gcash

[timid plank]

I would like a feature suno support true multi-persona collaboration inside a single generation. That would be a good community is generating good songs but lack that feature is hard I thing to create collabration

[minor pine]

Suggestion: Push this feature more. It shouldn't be buried under the more section at the bottom left of the page where no one sees it.

[indigo ginkgo]

"I'm tired boss" - me after seeing yet another update from Suno that limits visibility...

Time after time things just happen to us, the people that take Suno seriously, that spend most of our time here and I'm tired. Truly.

Why would you decrease the size of the profile banner, decrease the amount of songs I can pin, decrease the size of the song images I took ages to make and curate. The videos I put effort in making, why can't I pick the order, why can't the order stay consistent?

I say this as someone who has been on Suno's side for so long, who's know for being nice, a proponent of the AI music space and looking at the positive side, but how can it be that UI-wise you truly haven't made a good decision over the last year and a half and every single UI change has resulted in a noticable and proof-able drop in engagement on our music. (I literally studied UX-design next to my game design degree) and I cannot see it, so I will honestly ask: what are these UI changes meant to achieve and why should we be excited about it?

In what world does the new page look better than the old and how does it give more attention to our songs, what problem does it solve. Please help me understand?

I love you guys, I love suno, I love what it allows me to do, but this has to stop, because it is taking away all the joy, when your music is no longer being discovered and listened to.

[compact hollow]

Teach Suno music theory so musicians can enter exact instructions like key, scale, BPM, time signature, and chord progression, and have Suno actually follow them. If I type F harmonic minor at 110 BPM, it should do that exactly. This would make Suno much more useful for real producers and songwriters who build tracks around the vocals it creates.

[hollow night]

a cut time feature, like in ableton

[hollow night]

a merge waveforms function would be amazing

like all the bits and pieces in a channel becoming one waveform rather

[solid cargo]

its because they want to unify their mobile app with the Web interface. Instead of building on a good foundation they want to push you to use the app.2

[ocean forum]

I remember when the front page was so much more interesting. There is nothing on it for me now

[soft bridge]

Please continue to make prompt adherence amazing, particularly with the lyric box. In fact, I think that stronger prompt adherence is more important than size of the training dataset for the upcoming model. Please consider allowing advanced users to use seeds.

[indigo ginkgo]

If it's demonstrably worse for everyone, being consistent isn't a good argument to me.

[tepid laurel]

make the song list / library have a grid view that makes it as compact as possible so i can find my songs quicker.

[tepid laurel]

man i literally blow through the entire 12k in 8 days..

[tight summit]

After the “brilliant” decision to hide the number of credits, I realized everything I needed to know about those wonderful marketers. Looks like the WB marketing integration has officially begun.🤪

[rare grove]

You're 100% right. I won't publish any more. What is the point of being virtually invisible both on your own page and on the main page?😠😥

I will probably let my subscription expire as well; unfortunately, it's an annual subscription because I really loved Suno.

[rare grove]

You're absolutely right!!!

[teal mantle]

Came here to complain about this same thing the minute I saw it. Pinning songs effectively buries the previous pins. You may as well only pin one song so the most recents are still visible without the scroll. Or don't pin anything at all.

It just seems pointless saying any of this at this point, I don't know why I bother.

Suno has a lot of interesting things going for it that make it, potentially, more than just a music player. But it feels like the UI changes every week. You can't build anything on a platform made of sand.

[tropic pumice]

The UI gets worse and worse. And i can't remember when something like that ever worked on a PC interface. Remember windows 8... And it's even worse here cause as you say, it kills visibility. This needs more attention. Nobody i've spoken too likes these changes. It just feels like the UI team is bored at this point and has to do something. Never seen this many UI changes...

[vocal dune]

I am a massive fan of v5.5 and upgraded to the studio plan. My biggest problem with it is that songs (even the background music) have an "essy" sound to them that they didnt have in v5.

This problem is consistent with multiple professionally recorded and isolated vocal samples, even if I highly process them with noise removal and de-essing. It's also a problem in 5.5 even if I dont use my voice at all.

Also I would love it if you would add a feature that allows us to deem another user a collaborator and use their custom voice in your songs, especially duets. I wouldnt want this to just be a toggle where your voice is public and anyone with your username can now use it, but a permission you can grant to specific users.

Also duets with your own voices like pick voice a and voice b and then in the lyrics simply insert in brackets [voice a] where you want that voice to sing the next part.

[plush talon]

I myself am totally blind and have been trying to get in touch with the suno developers about major accessibility issues for students like myself studying audio engineering and do not know who to contact. I have even sent them letters threw their legal channel threw postal mail with signature required and my messages about violations to the Americans with Disabilities Act (ADA) have not been addressed at all so does anyone here know who I can write to to get these major issues resolved?

I myself was wondering as I am totally blind and have tried to contact the suno developers about major violations to the Americans with Disabilities Act (ADA) and have even sent them postal mail with no luck so is there a way to contact the developers directly about this as for us blind students studying audio engineering these issues are major?

[plush talon]

I have also noticed that my custom voice does not sound like my real voice at all no matter what I do like create a custom model with just my vocal dry tracks uploaded or use a high quality microphone to do the voice custom feature the voice regardless sounds like some strange man and not like how I normally sing or sound at all.

[steep dove]

Suggestion: how about actually paying out the contest winnings on ALL the contests not just some of the contests. This one still hasnt been paid and several contests have come out since...

[sullen valve]

I would love to try this use your own voice feature, but suno studio won’t pick up my voice recording, I’ve tried my interface, my usp microphone, my default microphone, and nothing picks up my verification recording?!?!?! What am I doing wrong here PLEASE HELP

[sage bay]

Please add a Unplayed Songs filter to the filters.

[pearl carbon]

@daring tusk @grave salmon

[cursive sinew]

Prepare to be disappointed. Their goal is mobile device engagement metric (Tiktok, IG, Shorts) just see how the phone app is now, it's 99% HOOKs, nothing else and doom scrolling.

All these changes on the larger scale is designed to take control away from the users, and let them curate everything you see and do.
This is typical late stage corporate moves to push a product out to the masses even harder.

The community at large, will not get an answer to any of this. Cause the bottom line is, they don't care.

[frank junco]

Hi, I don't know if it's already been suggested, but can you make a Playlist searchable please and also add an option to remove a song from a playlist like adding it to one ?!? Thank you, it would be very helpful when you update a song and want to replace it in a playlist!

[sleek abyss]

For awhile, when you'd do "crop" it would automatically create an upload of the same song with the auto-detected lyrics and style... THIS WAS SUPER HELPFUL. Make it a feature.

[small sand]

Just adding my voice to those requesting a revert of the new profile page design for the website - it's a regression on every front, almost all the information someone needs when they land on your profile page is now buried:

• the profile description is buried at the bottom
• song genre tags are truncated
• half as many pinned songs and users have to click through to see them all so they can't just gravitate towards one that looks most interesting to them
• thumbnail/poster images for other songs and playlists are so small they don't convey any sense of the song's style or identity
• the pinned carousel is prompting me for a caption for tracks that already have one, and if I add a new one it is not remembered
• so much empty, wasted space...

[tropic pumice]

That stupid hook nonesense needs to be deleted.
We paid them all the time for an A.I. Music tool. To create songs and that's it. Not for a second tiktok where they can make much more and say fck off to everyone that helped them so far.

[safe ingot]

Why did we ever ditch that layout, it is still so beautiful and functional compared to what we have now...

[teal mantle]

It looks like they're trying to unify the mobile and desktop UI, the problem being that the mobile UI has always sucked and they're making the desktop more like the crappy mobile UI

[dawn flint]

New Profile UI:

• Profile page puts all Pinned songs into a carousel.. so they are effectively now more obfuscated than non-pinned songs = more useless space doing useless things
• removal of 'Section Ordering'
• No change/but no fix: Old songs with 'video covers' play on mouse-hover - new songs don't have this cool feature

'New' Workspace UI:

• what happened to the credits counter?

v5.5:

• EDM genres have an inherent 'Spectral Haze' that sounds like a Vocoder is placed on the Master Bus
  • Songs sound like they never leave the root note, and melody sounds like it operates on harmonics instead of actual notes.
  • Rock/Soul/non-EDM don't seem to be affected.
  • ChatGPT seemed to be aware of these 5.5 model tendencies & suggested to add spectral Blocks to direct the model: which effectively chops 500+ chars for EACH song with the apparent new requirement for Spectral block directives to carve out something that prob shouldn't be there inherently.

Overall UI:

• Nearly convinced every software/web UI developer seems to be hellbent on catering to keyboardless-mouseless mental midgets - in some Tavistock MKUltra operation to dumb users down - as UI components munch obscene amounts of UI realestate with padding & monstrocity-sized UI elements to cater to tablets & phones for users that can't seem to get off the couch long enough to do something productive (b/c these are the users that are really going to sustain a platform by their sheer seriousness).
• any sight of 9:16 aspect ratio evokes a visceral recoil with me asking 'why am I here?': song covers/hooks comes to mind. Humanity spent 50years figuring out that 16:9 is optimal viewport for beings with 2 eyes. 9:16 is for the recessive participants of the planet dedicated to frying their body parts with 5G radiation while doomscrolling until greymatter oozes out of one-or-more-orifices.
• Can only assume some Global UX Steering Group pushes these ideas as 'Next Gen', but = Digital Green Agenda no one wants.

[glass barn]

Teach Suno bot about reverb and more stereo effects

[undone nacelle]

Is there a way we could implement an "analysis mode" on songs we make? I know Suno will give you a description of a song you upload, but for someone who doesn't know crap about music theory, that kind of information is helpful for learning how to make songs in that "taste" and play around with it.

[dawn flint]

BTW: Overall, LOVE the platform & the product - so commentary is def relative. The reality of slapping some text into a few boxes & clicking a button to conjure a song from the ether remotely approximating my intent & mixable with effects - is SciFi-level "No Way!". Audio fidelity & prompt adherence improves with each model gen, + STEM'ing & muxing, + multi-tracker(tho rudimentary), + DAVE on Discord deftly curating our tantrums - so there's much to be appreciative of.

[spiral bronze]

They don't even pay people who do paid surveys. I've been waiting for a month. Second time this has happened. They'll gladly take your money to fund their TikTok at Home™ R&D bender though.

[rare grove]

Bumping this!

[odd wraith]

there should be a feature to add notes onto a song so you know why you never published, what was, what needs fixing, really anything. it would save a lot of time and effort for people that have to get up and go do other things and cant work a song to completion from start to finish in a single go.

[jagged pier]

I would like to recommend that they should be a way to be able to keep the prompt enhancer when my taste is active, as to not to have to have to deactivate it each time someone like to use the style prompt enhancer.

[bold fulcrum]

don't know where to write it, but when we upload a hook, when we have a moderated video, we don't know where there is a problem, just a popup or a message saying what was wrong in the hook would be great to edit it. Right now we're going blind... Thank you, Apotiker.

[plush talon]

I also notice that if I create a cover of a song I am unable to download it or extract stems for the cover just the original song if I try to do so to the cover and if I try to delete the cover the original song gets deleted so I have to go in the trash and recover it so can you please fix this behavier and whenI look at the song name it is the song name of the original song which I guess makes sense?

[misty scaffold]

Thanks for let us making songs with video covers just to change everything so nobody will ever see those. not in trending, not in style lists, not even on the profile page. good job just make everyones experience worse every update. you can make the best models but this will make suno disappear as soon as others will catch up with the models.

[knotty rain]

First, something the team and/or model have really gotten right: I used to have a devil of a time getting certain instruments, like baritone sax or clavinet, to show up for improv solos, but more recent attempts in the past few months have given me some really nice ones on not only those two instruments but also more unusual and even obscure ones like banjo, harp, bassoon, biwa, and sitar.

I'd love to get even more, like bass and even contrabass flute, and perhaps even tenor sarrusophone (though that one would definitely be a very tall order). But I'm pleased with this.

All that said, I still have times when I'll specify something like "tenor sax solo" and get an electric guitar instead, or vice versa.

Also, I'd like to have a toggle with the "Cover" option to tell the model to not add extra measures.

And completely separately, I'd like the option to be notified when someone adds one of my songs to a playlist, or does a cover. In both cases, I'd like to be able to thank them.

[ivory frost]

Sorry, but I don't like this - where is the 'x' to get rid of that 'quest' window - It all feels so crammed already in there and now I can't get this collapsable out of the way - it makes me feel claustrophobic, honestly-
I don't care for free credits - if I want credits, I buy credits, simple - don't force that stuff on people please.

There seems to be such a big discrepancy: you are almost wasting real estate on the profile page and on the other hadn in the 'creative creation section' you restrcict and enclose, that makes no sense, sorry - it gives anxiety lvl 9000.

Love suno - but not so happy with recent changes concerning artefacts and spacial decisions, sorry.

i know you can hit the little arrow and it collapses down, but by god - get rid of it alltogether please - or give me the option to

[rare grove]

I don't just want to complain, so here's a constructive suggestion: How about, instead of the pointless carousel, giving users the option to present various songs by other artists that they currently like? I wouldn't mind if I couldn't pin my own songs anymore.🙂

[teal shore]

i think "Audio Influence" should be replaced with 2 new sliders: "Instrumental Influence" and "Vocal Influence", sometimes you only need to copy one, but not the other

[wispy sundial]

I think there should be an option for how many voices you want in your song - example if you want a group - you could pick "Group/quadruple voices" and if its a duet then "Duet" and if its solo then "Solo" and in the lryics section you could describe the voices in detail

[slow walrus]

SUNO HOW Embarrassing you want to get?????
IS THIS A JOKE? You really mean that serious Suno? @grave salmon @daring tusk @tepid trench Whoever is able to give a statement - a statement would be appreciated. I'm maximum confused.
Suno's moderation filter has always been the worst joke you could imagine, but this lets the barrel roll over completely. Blocking a music genre is a new low of maximum embarassment

Maybe you need a learning lesson?

https://www.metal-hammer.de/genres/hardcore/

Next step is blocking the genre in the Style box so that I cant generate this genre anymore? Is this gonna be happen too?

What is up with your site? It gets worse and worse. Let alone the terrible new profile design

[wanton cove]

i use my own prompt had no problem unyill the upgrade now every time i create ambient or meditive music suno puts a deep bass drone which actully makes the window rattle and drown all other instruments i suggest you get it fixed pronto fed up of wasting credits

[neon hinge]

https://suno.com/s/mjR7ViUKCiMUxHC5 I just did the exact same prompt with no problem, Tried hardcore with a couple of styles as well, all work. I would say the problem you're facing with it is an error or a glitch and that sometimes happens. Shouldn't, but it does.

[slow walrus]

its the style summary, not the prompt box

[neon hinge]

whoops, my bad. In that case, yes same problem

[slow walrus]

it started being a problem forever at profile page genre field
Now it did expand to style summary (why would you even add a censorship filter on a genre field is beyond me in the first place)
I wouldnt be surprised if that insane stupid filter lets me soon not generate the genre. Nothing surprises me anymore at this point with these UI devs and the even worse content moderation filter. I've nowhere seen a filter that broken and horrible. It even blocks in comments entirely normal german words. its crazy

And that it did now expand to Style summary, while I did report it several times for the profile page genre field shows me Suno simply doesnt care

And we probably also keep on this new profile design - despite absolutely everyone hating it

Sometimes I wonder why we report broken stuff and bad changes when they do what they want anyway. I'm sorry but at this point I'm more than frustrated about Suno's community care

And no statement whatsoever about the "hardcore" issue either. I reported it so often. Pure Silence.

[clever harness]

Easy fix. It'll go out on Monday.

[full bay]

come on guys! Ctrl-Z does not work in suno studio, basic stuff

[mystic sigil]

Ctrl-Z used to work - not sure what happened but very annoying

[safe ingot]

Too hardcore for Suno

[little forum]

Would be cool to get a desired Song Length function in the More Options section if possible.

[spiral bronze]

Considering I can't make instrumentals longer than 2 minutes with 5.5 but 4.5+ had no issues making 6 minute instrumentals, this is necessary.

[molten bluff]

Just curious if there was any kind of API access to Suno yet? It would be cool to integrate with my custom assistant as a skill to generate songs or sounds. Is it 'later' enough yet? 🙏 #💡┃suggestions-web message

There are also quite a few suspicious sites out there claiming to be Suno APIs and even using your name in the URL. By not having clear info about API offerings and not having this very common AI feature, it may open users to getting poached or scammed by one of these imitator services.

[distant sigil]

No APIs at all there’s been nothing updated since the video came out they don’t plan on it
https://discord.com/channels/1069381916492562582/1417458320708730880

[molten bluff]

Yikes, ok

[rare grove]

Unfortunately, this is still the case.

Reposted Seb's post

But it vanished
honi soit qui mal y pense😐

[slow walrus]

also to you: It is about the style summary and also profil genre field, NOT the prompt box!

[dull ridge]

I'm going with the duo function 2 saved voices on the same track

[sonic fern]

5.5 copyright detection is a joke. It's triggering on a song with literally 2 words lmao. Also triggering on songs that were written 300 years ago. Might consider fixing this aspect of Suno

[jagged widget]

Would be nice to bring back the trending playlists, and also have some way to filter out bots. Or cycle more often to the best of playlists.

[soft nest]

Mmm.. no, they proven time and time again they just stop updating them, the curated lists are also just the opinion of a few poeple and the community gave numerous feedback they don't want hand-picked lists anymore as they're unfair to like 99% of the others not picked. We need a better, improved trending system, pure and simple..

[jagged widget]

I agree, but that’s also part of it. Not all people who make ai music are equally skilled and or talented. That’s just life in general in all aspects of everything. Of in that 99% it should at least to me be a motivational thing to learn and want to be better. A simple fix for it, would be a set cooldown period after hitting a list. You come off and it’s a cooldown of a set time. Allows for rotation and movement instead of it always being the same people etc.

[soft nest]

That is true. Another problem however is that they removed the whole regional trending system too, which was a really good way for beginners or intermediate creators to get some visibility and recognition. That's also gone, along with playlists, so currently.. there's simply no way for people to find your songs -> zero visibility. The latest changes with the UI made this even worse.
@daring tusk and @grave salmon I really hope Suno reconsiders these recent changes...

[astral sigil]

yeah suno visibility now is simply hooks, if i make hooks i get like 10 times extra plays

if i dont make them? 100 plays

and my tracks used to average 1k plays . it’s all scaled down not inna good way

i mean i feel that instead of getting some fake engagement detection, they think these are solutions, sure you cut out the fake engagement but you cut out MOST ENGAGEMNT

fake or otherwise

same with certain users consistently making fake accounts to harrass other users who why not proper bans? they can’t litter your front page as easily when you simply ban their ISP.

[rare grove]

You're absolutely right!

@all: Go and take a look on Seb's post about the problem!

[zinc raven]

Copyright detection is triggering on anything and everything today. I have nothing but instrumental metatags, created from scratch, and it keeps triggering no matter how I rewrite it. Even when I tried to extend from a rendered song with just the outro, it triggers on that. Something is very wrong.

[clever harness]

Got an example?

[zinc raven]

I do, but I don't know what parts are doing it and it's long. It's happening on almost every variation I try. I'd share one but its way over 2k chars just for the lyric field stuff. Don't think I could even link a google docs upload in channel here. @joeyy was the one with the two word trigger...

[slow nimbus]

where's the persona tag in home page song list ? why you remove it? explain to me?

[tardy anvil]

what is with the new profile layout? square thumbs cutting off the vertical thumbnails, pins are only 5 now? and only 1 shows at a time (so the rest feel very hidden compared to just recent songs)...

also holy wasted space.

[tepid laurel]

maybe its a sign that all cover art will be square soon -- as it shouldve been all along

[tepid laurel]

I'd love for it to show what custom model was used, not just [custom]. which custom? I need to know

[astral sigil]

one last point about hooks and visibility, i personally quite enjoy making them, i am not a consumer of short video format, but what feels very very strange is that in order to get visibility on a MUSIC app is that you gotta make videos

there is a big difference when hooks was on top of “normal” algorithmic visibility, one who’s not into SVF can just ignore the hooks. now there’s no user playlists, no genre carousels, no front page to speak of, and the only visibility is from suno bot generated playlists which are so slow in updating there’s like 2 or less 5.5 songs on any of them, those lists also missing like some very basic large genre

i’ve literally stopped looking on suno, i also likely stopped curating any playlists on suno. the website is 0 fun now, if you’re trying to get people to see suno as an ecosystem, this is NOT the way. i mean in a way it’s saving me a lot of distraction and am now pretty much solely using suno as a tool since the last two UI chamges.

[teal shore]

if they don't make some relevant changes to visibility and necessary UI updates and feature fixes by the end of this year, i'm going to unpublish all my songs and distribute them to external platforms only instead, because it feels like that is going to be the better option for all of us in the long run. there is literally no reason at this point, for anyone to post songs on suno because they'll only ever be seen in the servers we post them in 😮‍💨 i feel like suno is taking notes from highguard's dev team

[young zenith]

Okay, hold on for a second — I have a suggestion for the pinned songs. It’s just a suggestion. I put this together in Claude, and I’d really love it if pinned songs could behave something like this :) 😎

[teal shore]

they should just revert the changes they made, probs cheaper and easier, but regardless your version looks a lot better than their minimalistic approach

[tardy anvil]

D: it WAS square from like v2 - v3.5

[vocal vapor]

The ability to rename these would be a god send, just saying.

[dire dune]

On the website when going through playlist, can we have an option to move any track to the top of the list. instead of having to drag it up. idk anyone else would like that to.

[tribal pewter]

Yes, since the recent layout changes, I have also been missing the option to view trending songs by language to see what is popular in my own language, Dutch, at the moment. It was also my only way as a small creator to gain attention among the entire Dutch-speaking world. The disappearance of the trending playlist also makes it harder for us as creators to be found, so in addition to rolling back the layout changes for the profile, home, and explore pages, I am also calling for the return of trending by language and the trending playlist options.

[cursive sinew]

If you go check the trending lists now (any language except English) you will see non are updated, only Global and English are updated with the exact same contest entry songs. The idea of trending is over. It is now a curated "Neon chains" experience. This is regressive modernism in every sense possible. Catering to mobile device metrics and inorganic user growth and song plays. (Sorry but I do not believe the current remix contest user has 3.5mil plays organically when most of the songs they have sit at sub 100 likes with 50 - 60k plays on each song)

[soft bridge]

Could we have the option to filter our libraries, and workspaces by song SIZE? We have Sample now, and finding some of those old 7:59 hallucinations with gems of pretty stuff can be challenging to find. I mention workspaces as well because, some of us would end up sorting through Sounds and stems as well.

[solar cliff]

If a user is blocked, please have some way to make sure they don't appear in any of my random plays coming up while checking iut other stuff - unliking a track and blocking someone should be enough to not have to listen to poorly done slop tracks again and again. Hell, unliking a track alone should stop it from reappearing again. I only resorted to blocking them since unliking seemed to fail for stopping the track from playing again and again and again

[teal shore]

i honestly think it's hilarious and sad suno uses bots to boost our stats, it's like clout chasing energy and fear of not being the best. instead of trying to be the best they should be trying to be functional and optimal, but there we go, they are going backwards not forward

[clever harness]

Why do you believe this is true?

[wild nest]

when are we going to get more control over the length of our songs especially when doing an instrumental? I don't want to be forced to use the wonky extend song and then have to try and piece the two together which doesn't work very well. I don't get the result I want and the extended song rarely sounds like the original. add back the Instrumental button and put in a slider when selected to influence the length of the song like you do with the weirdness and style sliders. I don't like it that Suno defaults to 3-4 minute songs.

[cursive sinew]

I dunno.. check the new remix content creator? 23 songs, 3.8MILLION plays and 8.8k likes? This is not organic, this is artificial boosting and you know it.

[clever harness]

Which account?

Oh you mean Bloomer?

[sharp meadow]

Features???? Like gen collaborations? Shows up on both artists profiles, Would be cool

Maybe like something I’ve said before, like a live collaboration system

Real time gen modification and communication in like an independent part of the site?

Maybe a matchmaking type thing built in too even, matchmakes you with another user that gens the same general music as you to do that collaboration with based on the common tags on your profile

[tight ore]

Dear colleagues, good afternoon.
Before writing my suggestion, I want to thank you for creating such a wonderful tool as Suno.
Now it's possible to "revive" my poems that were written earlier and give them a sound and make a song - it's very cool!
Now the suggestion for improvement itself.
At the moment, Suno does not know how to divide parts of the text into voices so that part of the text is sung by a man and part of the text is sung by a woman.
For example:
[Verse 1][Male voice]
one one one
[Chorus][Male voice]
two two two
[Verse 3][Female voice]
three three three
[Chorus][Female voice]
four four four
When generating, Suno does not understand these tags and may have a chorus where a man sings two two two - a woman's voice sings, and vice versa, a woman's verse three three three - a man sings. And you have to do more than a hundred generations to get a work that is approximately similar in meaning.
At the same time, specifying tags both in the "duet" style and specifying tags in the text itself [female voice][male voice] is currently not working.
Please consider creating a future model where it will be possible to divide into voices and indicate that this part of the text is sung by a woman and the other part of the text is sung by a man.

[teal shore]

because of the way they advertised the new model by saying it's the best, when it's barely different from v5. it's like we went from v5.4 broken to v5.5 broken, and it feels like their sales department desperately boasted it for new members. no idea why, maybe fear of losing people because the very controversial changes made. especially when there are better and more efficient ways to regulate music, now we are all hidden like ghosts, never to be seen or heard.

like if you wanna stop people spamming songs so there is less output, make a limit on publishings, like 1 a week.

if you wanna stop people trending, delete trending. its as simple as that, instead they turned the entire site into a bot-infested graveyard.

the site looked great the way it was, they just needed to focus on other things, real things like user feedback.

At this point, 4.5 is better than 5.5, it just has slightly lower quality, but it seems like there is a lot more diversity in making songs on older models than the new one, the new one is just clearer, and that's all.

[static python]

I tried the number 1 to 82 and got a copyright warning. That's BS of the highest order

[slow nimbus]

Bring back the persona tag!!!!!!!!!!!!!!!!!!!

[static python]

And a list of countries Alphabetically was blocked for Niger, which is a very different thing than the other word which is bad. When I spelled it wrong I got hit with another copyright warning.

[knotty rain]

I've actually had a fair bit of success with stuff like this:

[Verse 1 - Male]
This is the first verse
I guess it could be worse

[Verse 2 - Female]
I've got verse two
I'll share the chorus with you

[Chorus - Both]
Now we've got the chorus
And that's about it for us

And in the Style prompt I include "one male and one female singer."

Occasionally I do still get the wrong singer on something, but opening the Editor and regenerating that particular verse usually gets it straightened out.

[loud kite]

5.5 has an awful sound! Too bad. I hope it is still "beta" and it can be improved!

[spiral bronze]

Well presumably 5.5 is not quite a v6, so we can expect v6 to be better. I have noticed the terrible noise in v5.5 isn't quite as bad now as it was two weeks ago. Still there, definitely, but less obnoxious. It doesn't do anything that v5 didn't already do, it just has better more stable low-end.

<@&1082792303976001547>

[acoustic dragon]

Hey Suno team, As artists start generating hundreds (or thousands) of tracks, our project spaces quickly become cluttered. Many of these generations are worth keeping — we just don’t need them cluttering up our active workspaces anymore.Proposed feature:
Add a simple per-generation toggle called “Auto-Archive” (on/off). When enabled on a generation, it would automatically move that track to a dedicated Master Library (or Archive folder) after a user-set period of time — e.g., 7 days, 14 days, 30 days, or a custom number of days.This would:Keep our active project spaces clean and focused
Preserve every generation we don’t want to delete
Automatically organize our long-term collection without manual effort

It could even be extended later to work on entire folders/projects or include smart filters (e.g., “auto-archive anything not favorited or not used in a project”).I realize you might already have work happening in this area, but this felt like a really practical quality-of-life win for power users who are starting to build massive libraries.Would love to hear your thoughts or if something similar is already on the roadmap!Thanks as always for building such an incredible tool

[loud kite]

Please change the order in the workspaces list when you want to move a song from sorted by "creation date" to "recently used". Because we use to move multiple files (but not at once) in the same directory and scrolling every time to find the right workspace it is a stupid waste of time. Having the last workspace used to copy smth into at the top would make the job faster !

And one cannot still filter the "unlistened songs"... this is such a basic feature and after 2 years is still missing despite many suggestions.

[astral sigil]

it’s been appearing that way forever, the contests are all bloated, even prior to those all those 3.5/4.5-all free slop getting clearly botted (hello Lust in it / Muse in the wires and co)

[shrewd aspen]

Hello, good morning. I want to recover an account I have linked to Discord, but I accidentally entered the wrong date of birth and my account got locked. Is there any way to recover it, since I have two songs on that account?

<@&1082792303976001547> help me

[teal shore]

upon release of next model, instead of telling everyone it's the best, they should tell us what changed in it so we can use it efficiently and it wont feel as broken as the previous, like a mini-changelog will do, nothing complex, "this model requires less prompting than usual", "this model is better for personas", etc

[modern summit]

i completely agree with that

[teal shore]

here's an idea, scrap trending and replace it with a global spotlight like soundcloud has which is for paying customers only, so that it's not competitive and has balance, people are seen but not spammed, everyone has a chance to be heard, you get 1-4 slots which you can change freely, similar to pins but on the homepage instead of profiles, so like it will say "Spotlight" then below the portraits of all the artists who have decided to spotlight their work and when you click the portrait it shows you a mini dropdown menu of your spotlighted songs

[willow blaze]

Could anyone just kindly ping me once all the bugs are fixed?

[clever harness]

How is that different than the playlists or best of <x> in your mind? Those aren't competitive. Everyone has a chance to be on them.

[feral fable]

please give paid users the ability to block their content to guest and unsigned in accounts. there is no point in giving people the ability to block user accounts if those accounts can then sign out and view the content anyway . Thank you

[frigid bough]

The new version of Suno is trying to make songs shorter, not longer... I've spent over 1000 credits and it hasn't made me a single song longer than 3 minutes... And make sure you don't use "long song" in the prompt because it makes them even shorter...

My suggestion is that Suno stop wasting tokens on creating song titles and images to create songs of 3 minutes or more.

[wind abyss]

@daring tusk : songs that have the upgrade tag on, can not be liked in preview mode, they can only be like by the creator in the library section, if it in preview mode it can not be liked by any one, creator or listener .:) don't know if this is a bug, or a preference?

https://suno.com/song/a9d125da-7ba0-459c-835a-8c33fb863a45

[teal shore]

because its the first thing everyone sees all the time, even if it was a page similar to how the genres are displayed for in the genre-based playlists, trending is 1 slot everyone tries to cheese their way into, but you choose to be on this (possibly even be able to be curated through it by users who enjoy making radio stations and playlists). it would not be much work for suno because once it's done the users do the rest by uploading reposting etc

[shadow summit]

If you changed the remastering to only create one new version at half the credit cost, it would be much more useful? I can understand Suno creating two versions for a new song or a cover, as those versions are typically quite different from each other, but what's the point of creating two nearly identical versions every time I want to remaster a song? At least, have the option to create one subtle and one normal remaster version simultaneously if you really want to keep it at two generations.

[safe ingot]

Some apparently more than others

Also it kinda seems like they are suggesting having users pay to be spotlighted. Which... no

[clever harness]

In an ideal world the homepage would just be every song by everyone 24/7.

[spiral bronze]

What would also be nice is besides a remastering strength, being able to tell the remaster algorithm what is wrong with a track or if it simply needs more X, Y, or Z.

[dusty relic]

Any chance an LLM aspect could be added to the copyright detector to override all of the false positive it gives? Right now the lyrical copyright flagger is a fairly blunt instrument, but with an LLM or something it would be able to verify if things are actually copyrighted or just false flagging things

[near pelican]

Please add option to export "Lyrics time sheet". It would be simple text file where for each line of text timestamps would be added. It should be pretty straightforward to implement but would immensely help with video creations, especially with automated tools

[golden moth]

The censorship of links needs to be relaxed alot

We should be allowed to share released music and websites we made to host them

Stop req only suno links stop it stop it stop it

Remive the pointless restrictions from chats its just control and suno links share meta tags stop it stop it stop it

[visual osprey]

Please just bring back personas that I can use with songs that have both lyrics and instrumentals, its the only way to get motifs and reprisals and that is now gone. Just a toggle when making a new voice that doesn't strip away the damn instruments, PLEASE.

[knotty rain]

All the bugs?

What's your life expectancy?

[grim swift]

would be really cool if we could manage multiple different artists in a single workspace so i don’t have to have multiple plans and accounts

[slow nimbus]

We actually already have that feature, called 'Persona.' But now they've just turned it into 'Voice' , the purpose has changed now sadly. They (Shameful SUNO MF product team) even removed the persona tag on the homepage and profile page song cards.

And one of mf here told me to stop yelling, i mean dude, they don't hear anything

[grim swift]

persona is not quite what I’m referring to. songs still show under a single artist name in library and when published. unless im missing something

[slow nimbus]

yeah, i mean, Persona should be functioned like you refering to

[slow nimbus]

But now they are gone,
i have same idea as you, but before 'Persona' evolves becoming that idea, they already scraped it.

[grim swift]

different than persona’s

[cyan peak]

Ooh can you bring back the Showcase bit on the Profile Pages where you can see about 7 songs at once and not the clicky thing (which you can't add captions to). Also the old Suno Homepage was sooo cool where you could see everything at once, the new songs, and all the different Genres in one page.

It's hard to find new people or songs - just the ones that Suno like (which is cool but so many cool people get missed).

[teal shore]

because those playlists are hidden behind pages and aren't curated very well, it feels like the person/people curating them only like what they like and that's what everyone sees, which is very low quality in comparison to somewhere like spotify, where there are thousands of curators compiling and sharing playlists of all genres not just specific "popular" genres

[clever harness]

Makes sense to me

[rare grove]

Seb wrote an excellent post about the new UI and the problems. Already a lot of my friends left Suno and I won't publish any more.😐

[wispy hemlock]

The expansions still ignore everything you input. If you want to modify only the ending, Suno doesn't care; it finishes the lyrics and starts them again to create a 4-minute expansion. It doesn't matter if you put "[end]" in the song lyrics; Suno will restart the lyrics 90% of the time. What a ridiculous way to waste credits.

And you've limited the number of pinned songs to 5 again instead of 10... and it's touch-based too... it's really hard to see all 5 pinned songs on a PC.

[tight summit]

I think I'll support you in this endeavor, my friend. I absolutely don't see the point in posting on this platform.

[trim flume]

I ran into this exact problem today. i wanted to try a unique song with unique solos for 2 guitarist, drummer, and bassist. Suno cant really track that many unique instruments sounds very well. I got excited when reading about the Extend song option. I thought I could split my song into 3 parts where each part would let two instruments shine. Turns out, the extended part of the song gets turned into generic music and doesnt extend the original. This option really feels mislabeled. Wasted about 250 paid credits experimenting with it as I already used my monthly Pro alotment. 🙁

[rare grove]

Thank you so much, my dear friend! Unfortunately, there's zero feedback and no understanding from Suno, even though everyone thinks the new UI is terrible. You can see it in the new songs too; activity has dropped significantly, and many people aren't publishing anything anymore. Newcomers to Suno don't know where to start. In my first year, I listened to a lot of songs from well-known artists in pinned playlists and tried out their prompts before I started publishing myself in my second year. Funnily enough, one of my really old songs is even in one of the new static playlists. I'd gladly do without it if they went back to the old UI...🙄😞

[tight summit]

Yes. Unfortunately... The interface has been completely redesigned for mobile devices, but the problem is that it's terrible on mobile devices too. 💩

[rare grove]

You're 100% right. I mainly use Suno on my phone and then via the browser (since the app isn't very good). I'm afraid they're aiming for something like TikTok. Maybe they simply want a different user group; whether this new group will all have the interest and financial means for a subscription remains to be seen...

[wispy hemlock]

It makes sense that they want to attract users who listen to Suno's music because users who create music don't usually waste time listening to other people's music.

Enough time is already wasted creating music without having to listen to other artists. If you ever have the time, perfect, but we have a life beyond Suno.
Creators support each other by liking each other's posts, but in most cases they don't listen to them, they just "like" them.

[rare grove]

I don't quite understand what you mean. With the new UI, basically everyone is invisible, meaning you can't find any special music (except for the few outdated playlists on Suno or if you know people from before).

This even applies to people who only give likes without listening to the music...🙄

[wispy hemlock]

My English isn't very good, but what I mean is that they're increasingly focusing on mobile devices, and I think that's to attract an audience that doesn't create music.
I agree that the current interface makes us more invisible (that too), and I'm noticing it a lot.

[rare grove]

Sorry, I didn't mean to offend you, and English isn't my native language either. What I meant was that the old UI offered a lot more music to listen to than it does now. It's basically just focused on the hooks, so it's kind of like TikTok 😐

[halcyon tangle]

Make it so we can loop playlists.

[halcyon tangle]

I agree, the site looks like trash now.

[teal shore]

not only that, but compared to circa 2024-2025, it's literally a graveyard. im in 20+ servers, and they were all booming and lively communities when i joined and now all of them are dead. the only part of suno that feels alive are the showcases. it makes me wonder what other communities like facebook are saying, because it's likely similar.

[formal pilot]

I'd even be open to other options I call getting songs "out of jail". They could show a button next to our songs to get the rights for it in a couple different ways.

1. get this song only, and maybe charge $2 each.
2. get that month's songs. Show us a list by month of the songs we created. Charge a month of Pro Plan to get all of those songs out of jail.

Note that this will likely get "forever free" people to give them some money. This won't unlock Studio or other Pro features.

On the database backend, they probably have a flag indicating whether the song was created under a Free/Paid Plan in order to track monetization rights. Instead of a Boolean, use an Integer. 0 = no rights (default for Free Plan), 1 = Pro or Premier Plan, 2 = purchased single song (out of jail), 3 = purchased a whole month (out of jail)

[dawn cargo]

📩 A Heartfelt Letter to the Suno Team: On Art and Emotional Connection

First of all, I would like to express my sincere gratitude to Suno. You have created a truly meaningful platform - a place where those of us with melodies and emotions in our hearts, yet no tools to express them, can finally paint our diverse musical pictures.

I still vividly remember the surprise of stumbling upon Suno back in version 3.5. At that moment, I was as overjoyed as a child who had finally found the long-awaited toy of their dreams. I have always considered v4&v4.5 to be a "Golden Age," where melodies could touch the soul and bring me to tears right from the very first notes of an Intro.

I understand and share the challenges Suno faces in balancing business growth with the pursuit of authentic musical values. However, if possible, I hope Suno could offer a choice - a separation between an area dedicated to "Hooks" and a "breathier" space for those who love complete musical experiences. It would be like having designated areas for smokers and non-smokers, allowing those with enough patience to fully enjoy the entire journey of a song without their emotions being fragmented.

What concerns me most at the moment is that the newer versions seem to be becoming too "safe" and generic. We miss the explosive, unconventional, and intense "touchpoints" of the past. When music becomes too predictable and "mass-produced," it gradually loses its deep connection with the creator.

I don’t want to be overly demanding, as I understand the difficulties of running such a platform. In fact, being gifted a free subscription isn't as important to me as the ability to create high-quality, emotionally resonant music. I am always more than willing to pay to support your development and to contribute sincere feedback.

All we truly hope for is a comfortable musical space where the joy of creation remains pure and complete.

Thank you for listening. Wishing you all a wonderful day!

[pale imp]

It would be great if you guys could add Beat boxing or Accapella also.. it always just turns to music though.

[modern summit]

acapella/vocal arangement works actually, but not when i want like only one voice for example. i've tried multiple attempts for a single solo vocal and it never worked

[teal shore]

why not just render a song, then click "get stems" this grabs the accapella for you, and if you make a beatboxing type beat, you can grab that too

[gentle gate]

why cant i extend an instrumental without having humming and vocals that are definitely NOT in the prompt

its driving me insane.

[pale imp]

Because it still makes everything into music still.. Beatboxing doesn't really work because it still sounds like music instruments.. Getting vocals isn't the same as accapella.

[teal shore]

i think the only way to do it would likely be to overload the "exclude styles" box so that it excludes all the guts it would naturally generate so that all you want to hear is what you've prompted 🤔, not sure if it'll actually work though

another way is to type your beatboxing as lyrics for example like "boots n cats n boots n cats" would be a 4x4 beat, enforced by beatboxing styled metatags and vocalization

[barren bough]

I am brand new here, so forgive me if this has been asked a 1000 times. I would be great if artists could here money. Have a buy me a coffee or donation button if someone really likes a song or Playlist

[arctic python]

For this reason I have to go back to V4 just to finish or end a track. Then often I am remastering it using "Subtle" in V5.5 so it keeps most of the tracks integrity, hopefully

[neat raven]

The handling of copyright material is partly really ridiculous. For example, the German text of the Lord's Prayer, which surely isn't copyrighted, is rejected because of copyrighted material.

[carmine kelp]

please make own-voice feature available on SEA especially Indonesia 🙏🏼

[pale imp]

I'll try... just something I've wanted to try with this stuff but always just sounds like music.

[young apex]

I have a suggestion that for me it would be very useful
Now the "Reuse Prompt" option replace the lyrics, the style and the title.
The most of time i just need to import the stile.
Having a option to select what the user want to replace would be great

[slow walrus]

yeah worst moderation filter ever. at least they fixed by now that you can add the genre "hardcore" in profile and style summary lol

but that it blocks totally normal german words is ridiculous

worst moderation filter on earth xD

just like the new profile UI design. All hate it, we keep it nonetheless .......

I have to give my followers new songs in discord now

because the profile page is so bad in presenting the songs that my followers dont even notice them. If you want the worst way of marketing your songs you use this new profile design of suno. its really the maximum wrong way how you should do it 😐

What was wrong with the previous design? nothing. Why we changed to this new disaster? Why we keep this disaster when everyone, absolutely everyone here hates it. I've not seen yet a single soul here who said - its great

[shell epoch]

How about letting us use our personas/voice models (however you wanna call them) in studio when we generate vocals or when we cover vocal lines we uploaded? Or make it so that when we cover/generate a single instrument it doesn't add a full band or orchestra behind it no matter what we put in prompt or exclude? Or simple things like putting a "select all" on trash when we want to delete the 50 empty stems it generates when we extract stems? Or maybe rename suno to wtfwwt (what the f.... were we thinking)? Worst and most difficult to use program ever made...

[knotty rain]

Formerly, when clicking on the empty part of a bar of a song within a Playlist, the song would come up in a right-side sidebar for easy playing without triggering the whole playlist (from that point forward, anyway). I'd kind of like to have that come back.

[upper wing]

when the popup that says new songs are hidden because of filters, it'd be nice if we could click on it to reset the filters to default!

[fallen frost]

When I want to record my voice and then make a song out of it, a completely different voice appears in the song. It's not even close to my original.

[violet sierra]

Did you have to ruin the side view album art? It's tiny and not vertical, also cropped from above very often, which ruins a lot of the videos. And the built-in video generator videos are even worse, they have terrible black bars from both sides, not fitting well at all.

[spiral bronze]

What do you mean by "side view" album art?

[rare grove]

I want to reiterate the absolutely unfortunate situation with the new UI. It should be obvious to everyone that placing the profile at the very bottom is simply a bad solution. The lack of customization options and this pointless carousel should also be changed. Furthermore, PC users are complaining about a "mobile phone look." I primarily use my phone, and the no-longer-recognizable thumbnails are incredibly annoying, as I put a lot of care into my photos and videos, and all that effort is currently wasted. It's also extremely frustrating that the music genre is no longer displayed in other users' song lists, preventing me from easily selecting which song I'm interested in listening to. I haven't found anyone who likes the new UI; it's completely demotivating and ruins the enjoyment of Suno. I myself will be switching from the Premium plan to Pro or Free, as I simply don't need 10,000 credits in the current situation. Unfortunately, my subscription runs until November, but I've already canceled it.😞

[jagged kayak]

SUGGESTION: The new "Featured Song" section on the profile is a major step backwards.

• You have half the songs
• it takes up so much blank space
• the captions you can add are too short
• It makes it harder - not easier - to see what songs I've missed by a user. I wind up going to all songs.
• It defeats the purpose of showcasing a song if you can't see them
• It doesn't look good.

I really see no positives.
What would be much better is a way to customize this section the way you want. Layout options, perhaps 1 song takes 2 tiles, and the rest are 1 tile. Just give the user 10 tiles and let them make what they want. But at the very least the old way was much better.

[kind grove]

I only use Suno now for generating content and a bit of post-processing. I’m not interested in any of the publishing features anymore. The cover display is completely off industry standards—why does it have to be in a vertical format? The profile page has become totally unattractive, and I have no way to fix it (no customization options). I create my cover images externally—gives me more control.

[silver harness]

Can someone add way to remove personas?

[tight summit]

Library -> Voices -> Person -> 3 dots -> move to trash
Or are you asking for a faster way?

[low lance]

I have a suggest for covers that we search for our favourite songs which are outside of Suno (i.e. Survivor's Eye of the Tiger) and that way we can cover our favourite songs without going through the long process of Uploading the audio to suno and then going to Remix/Edit, then going to cover. I feel like that would so much easier.

[clever harness]

Are you uploading copyrighted material today in order to cover it?

[low lance]

Only out of personal interest and don't worry burn it off of CDs I have so I'm not pirating Music

[clever harness]

Please don't do that, that's against ToS.

[low lance]

Okay but if my suggestion was to be added as a feature then would that still be against ToS or what will happen there

[clever harness]

There might be something interesting coming in the future that would not be against ToS.

[low lance]

Is that more or less connected to my suggestion?

Also I don't share those songs publicly

[clever harness]

Understood, but you'll want to avoid it in general as it puts your account at risk of being banned.

[low lance]

Okay, fair enough

I just originally thought it'd be a cool feature to have in the future but if something more or less is already in the works then cool I'll just be patient and wait

[spiral bronze]

I still think the ability to create "motifs" as a collection of melodies or patterns of textures and then using them for scoring purposes or to make reference to another song without covering it.
Forgive the weird example here, but what I mean by that second one is perfectly illustrated with the Sonic the Hedgehog 2006 soundtrack. Everybody knows the jingle behind the Green Hills Zone music from the very first Sonic game, but that motif was accidentally referenced by the musician when making Dreams of an Absolution even though the rest of the song is completely different.

I also think this would improve sampling, if it were possible to work multiple samples in and maybe even assign what you want to use that sample for. Say you could assign one sample as the drums or percussion, another sample as a backing instrument or vocals, another sample as a lead or a fill. Point is for them to really be samples, Suno ought to use them more intact and only really timewarp or pitch shift them. Well, if you say a sample should be a motif, then maybe that allows Suno to be more creative in its interpretation.

While I'm on that topic, that made me think about the mashup feature. I think it's probably my least favorite of the cover methods but I wasn't sure why until I thought about it, and I think what irks me about it is how it seems too rigid. I think if it also allowed up to four tracks (or whatever number) and could allow for medleys and not just cutting between two songs, it'd have more utility and be more interesting with instrumental tracks, since I think right now Inspiration does like 95% of the heavy lifting in my workflows primarily because I can take four of my best takes and weld them together. I think if mashup and sampling could do that then too then I'd use them much more.

[dire dune]

On Playlist, can we have a feature to auto sort. Like most recent at the top, Also a move track to top of list option.

[mighty creek]

Fix the playlists! They are so laggy, in the web!

@daring tusk optimize the playlists it is laggy for web

[teal shore]

pretty sure ripping songs off CD's is still pirating, it's only not pirating when you've asked the artist for permission to use a song

[slow walrus]

Regarding this I actually have a question:
What about music by Warner? Would it be still disallowed or does it change anything thanks to the deal with warner?

[clever harness]

TBH I don't know the details on that side of things.

[karmic sigil]

Remaster with Custom Model

[cosmic fractal]

So token expansion for lyrics I out would be nice I hit the limit on alot of songs and have to simplify most the time

The models clearly have the capability to go beyond the current limits I've seen 7-8 min generations that have full songs but missing lyrics cause them to either glitch and overlap and overwrite themselves or they end up restarting

Additionally not sure what rlhf is being used for the training process but maybe a drop or gpro maybe even my customer CCRL framework may help

CCRL introduces intentional entropy mapping to increase diversity rewards during training as well as the usual reward training

Interested to see about possible advancements that could be made to the training process ?

[oblique grail]

Hello Suno team,

I am a Pro user and I create angelcore music.
I love Suno but I have noticed that it struggles
with this genre. Here is what is missing:

1. Distorted synth leads — sharp, buzzing,
   high-pitched synth melodies that cut through
   the mix. This is the signature sound of angelcore.

2. Reese bass — deep, droning, heavy synth bass.
   Not club bass, not trap bass — dark atmospheric
   low-end rumble.

3. Airy plucks — light, ethereal plucked synth
   notes floating above the mix.

4. Angelcore as a recognized genre — please train
   the model on artists like akiaura, LONOWN,
   Crystal Castles, Øneheart, STM. This genre is
   growing fast on TikTok and your users need it.

When I type "distorted synth lead" or "reese bass"
Suno does not reproduce these sounds accurately.
Please improve this.

Thank you for your work.
Best regards,
STIKMEN

[cosmic fractal]

So as a pro user there is a custom sound generation option that allows you to actually generate the sound samples your actually requesting and describing ? 🤔

[limpid bolt]

Microsoft OAuth needs the select account prompt added. I am forced to log out of all of my other accounts just to log in to Suno.

[tidal galleon]

Feature Suggestion for Suno: Vocalist Recording Link

I would love to see Suno add a feature where creators can send a secure recording link to a vocalist, artist, or collaborator. The vocalist could open the link, listen to the reference track or section, record their vocals directly through the browser or mobile device, and then submit the take back to the original Suno creator.

This would be extremely useful for producers, songwriters, and concept creators who may not sing themselves but still want to collaborate with real vocalists. It would also solve a major remote-collaboration problem. Sometimes the artist or vocalist you are working with may be in another state or even another country, which makes it difficult to verify their voice locally or record them in person. A secure recording link would allow the creator to confirm that the voice belongs to the actual collaborator and collect an approved vocal take without needing everyone in the same studio.

A few helpful options could include:

• A private share link for vocalists
• The ability to upload lyrics or reference audio
• Browser-based vocal recording
• Multiple take submissions
• Automatic stem separation or vocal-only delivery
• Creator approval before the vocal is added to the project
• A voice verification step for remote artists
• A usage rights/permission checkbox before submission

This feature would make Suno even more powerful as a collaboration tool. A lot of users are not just trying to generate songs; they are trying to build real music concepts that can be developed with singers, artists, and producers. A vocalist recording link would help creators move from idea to demo to finished record much faster, while also making remote collaboration safer, cleaner, and easier to manage.

[west elk]

Suno can do all that stuff. Granted, it may not be working for you, but I routinely use all three of those techniques in my music.

Question: What's the point of "pinning" stuff in our profile if the profile page only shows the first 5 pinned items?

[oblique grail]

@west elk Hey Patrick, thanks for the reply!
I’m trying to make proper angelcore tracks, but I’m struggling with two things:
The bass — I can’t get that deep, warm, angelic bass sound that you use.
The leads — they’re missing that ethereal, floating, emotional quality.
Could you please share some prompt examples for both bass and leads that give that true angelcore vibe? Just a few good ones would help me a lot.
I’d really appreciate it, man. Thanks!

[spiral bronze]

It'd be nice to have a "Credit Reset Date" displayed with a "Days remaining" on the account page so you know how many days are left until your credits reset. I'm on an annual Premier plan and I upgraded while I was mid-monthly-Pro sub and paid the difference - this is back in the era when there were still people working for the Suno support email team and they actually answered emails - so I don't have anything that tells me directly when my credits will be reset.

Adding on to that it'd then be nice to be able to change when that date is to make it a better time for the user without unsubbing and resubbing.

[spiral bronze]

When you're setting up influence audio, if you have an audio influence percentage already set then it gets reset to default values when you add a voice. This has several times now resulted in me accidentally genning tracks with the wrong settings and having to waste credits to fix it. I still think that audio influence and voice influence should be completely separate sliders, but until then it'd be nice if it didn't touch a slider that's already set how we want it.

[pale imp]

Hey, it would be great if we could search for playlists! I have so many now, it would be make searching for an older playlist much easier.

[vital abyss]

add techno magic as a genre !! ... in fact add Techno in the #1154790975240155247

it's missing .

https://suno.com/search?type=public_song&q=techno+magic&sort=most_relevant

[haughty dome]

I think that MY TASTE could be influencing the complaint that many people have about all their songs having the same tune. It's hidden away under My Profile when you click your own avatar button in Chrome, desktop PC.So my suggestion is that it should be more prominent. BTW I don't understand what I've violated in this :

[hollow harness]

Would love it if we could message users that follow each other so we can collab, other than leaving comments which are public it makes it difficult to communicate with other users.

[spiral bronze]

It's against Suno's policy to follow your prompt, obviously.

[spiral bronze]

Also, I'm 99% sure that "My Taste" is not a masterprompt like LLMs have. It's just the prompt that gets injected when you hit the refine button on style prompts. If you don't hit refine, it doesn't do anything.

[clever harness]

Right, if you don't click the blue wand My Taste does nothing

[azure sapphire]

I would like to suggest that we totally disallow reusing lyrics if allow remixed is disabled because this opens up a can of worms on proprietary concerns.

And that if cover/remix is used, even private or unpublished versions of songs appear in original creator's view (for original creator's IP protection of course).

[teal shore]

what you're actually asking for is "add EDM" (Electronic Dance Music), which is the category in which "techno" is a genre within, like "trance, dubstep, hardhouse, drum & bass, etc" which would be a bit pointless because there are about 200 or so genres within "EDM" and the techno songs being produced would be pushed under all the genres people favor over it

they could however make a section labeled "EDM" but there is a lot of EDM prompted music on suno which is realistically rock music and other types which aren't dance music.

[vital abyss]

No. I want techno

[cosmic fractal]

new feature for song reveiw is broken doesnt render the audo logs only lyrics and tags tried to skip a few and see if that helped attempted a refresh of the page and even went as far as clearing the broswer cashe and everything for some reasong the audio just wont load in this format old format works fine this is a new reveiw process im assuming so id like to say it is 100 percent better and is clear on whats being asked i liek the tag on the bottom when selection of a or b ttracks

[woven burrow]

Lol he's making a song about worshipping Tencent and Xi Jinping

[cosmic fractal]

thats not my song lol but yeah i guess i didnt get to listen to see waht it sounded like this is the earn credits by reveiwing songs feature

[cinder jacinth]

Remix off by default is actually the right move because creators should control whether others can reuse their work. But hooks disappearing is different. Hooks were one of the few lightweight discovery tools creators had for getting attention and building an audience. If hooks are being removed or restricted, Suno should explain that clearly, because it changes engagement and growth on the platform. Creator protection is good, but removing discovery tools without replacing them makes it harder for smaller creators to get traction.

[teal shore]

well techno is just a genre, and to have that it must be categorized and if they do it, it wont change anything, infact, it'll make your life harder because you'll have to click "EDM" and then scroll through a list of genres and sub-genres until you get to techno and if you want that, good luck to you

[pallid dirge]

Can Suno put a post out on if they can't do MIDI imports for possible legal reasons. I hate saying this suggestion time and time again because I know this can be done.

[twilit lagoon]

Creating 2 sec tracks from 4 min covers tonight... wonderful.

[willow blaze]

My “mashups” are all coming out as 0:06. wonderful indeed.

[river tide]

We need a feature that allows us to zip and download all songs from a playlist.

[clever harness]

Do you have an example track?

[clever harness]

Do you have an example track?

[fallow zodiac]

Can you all add the allowance of more than 3 custom models? That's pretty limited, especially if people such as myself do around 10 different genres.

[clever harness]

You'll want to dial those sliders way back to probably < 60

The higher/lower the slider the more unstable the audio

Doing extremes = yolo mode.

Probably should just change the button when people go to the extremes...

[fierce token]

I'd like to request the option where it will generate a vocal track and music track and put them together. Currently the stems feature makes the tracks muffled so it's basically useless. It would be better to generate this as the song is made. Or at least give us that option if we desire, even if it's gunna cost more credits to do.

[tiny gulch]

Add the tag for J pop songs🙏🏻🙏🏻🙏🏻

[clever harness]

Could you expand on that?

[tiny gulch]

Japanese pop music

[clever harness]

Ha.. no I meant, where do you imagine that living in the UI?

[tiny gulch]

Um... Yes. It's the first time I've used the application, probably I didn't notice.

[ancient matrix]

is 6/8 time signature regressed in 5.5, i have a hard time getting 6/8 with 5.5, but dropping back to 5 i can get 6/8 again

[fallen ember]

There are a lot of channels and I don't know exactly where to put this, but what's going on with moderation? I don't even make spicy videos and my hooks keep getting rejected.

[clever harness]

What's your handle?

[fallen ember]

asifheard

[ocean bolt]

Hey Suno team! I don't know if this has ever suggested before - I'm sure it is at some extend.
As we can now use magic wand to do a spell to conjure new totally new styles and such to our tracks (that we never had even idea of) . Why don't create a "chat" to directly chat with AI to exchange ideas to change metatags (and maybe lyrics) that we want to our tracks? Would be even better if we could see all the metatags and such Suno adds to our tracks.

Uhh, English is not my first language so, please ask if you need clarification. 🤣

[fallen ember]

I would love to have conversations with the Suno AI too. For a variety of reasons 🙂

[blissful sluice]

The "Stems" Dilemma: Why Suno's Current Method Doesn't Work for Producers
The current "Stems" feature in Suno is unsuitable for professional post-processing in a DAW (Digital Audio Workstation). The reason lies in its methodology: Suno generates a finished stereo track and only then attempts to separate it back into individual tracks like vocals, bass, drums, and instruments using AI. This process is known as post-production separation.

The Core Problem: Post-Separation Instead of Native Generation
Imagine baking a cake and then being asked to separate the flour, eggs, and sugar back out. The result can only be messy. This is exactly what Suno is trying to do with its stem separation. The AI can no longer accurately assign frequencies shared by different instruments—such as the low-end of a kick drum and a bassline. The result is the well-known "muddy" tracks with artifacts and "bleeding," which make a clean mix impossible.

The Solution: True, Native Multitrack Generation
To get truly usable stems, the AI architecture needs to work differently from the ground up. The solution is native multitrack generation.

How It Works:
Instead of one finished track, the AI would calculate separate, perfectly synchronized audio tracks for each instrument from the very beginning. You can think of it as a virtual orchestra where each musician (each "sub-AI") records their own isolated track.

Advantage 1: Absolute Cleanliness: Since the tracks are separate from the start, there is no bleeding. A kick drum track contains 100% kick drum, and nothing else.

Advantage 2: Full Creative Control: Producers could specifically edit, replace, or regenerate individual tracks while the rest of the arrangement remains unchanged.

This method is more computationally intensive, but it's the only way to deliver the quality required for professional music production. Other platforms are already exploring this approach, where stems are created separately from the start to avoid quality loss.

Why Native Multitrack Generation is So Difficult: The Technical Hurdles
Switching from generating a final mix to natively creating individual tracks isn't a simple update; it requires a complete redesign of the AI's core architecture. Here are the biggest challenges involved:

Exponentially Higher Computational Cost:
Currently: The AI generates one complex audio stream.
Natively: The AI would need to compute 8, 12, or more separate but interconnected audio streams simultaneously. This isn't just 12 times the load; it's exponentially higher because each track must influence and remain coherent with all the others in real-time. Imagine conducting an entire orchestra instead of just listening to a soloist—the coordination overhead is immense.

The Coherence Problem:
Music is more than the sum of its parts. A bassline must lock in perfectly with the kick drum; a melody must harmonize with the chords. With native generation, all the "sub-AIs" (responsible for drums, bass, vocals, etc.) must constantly communicate to create a musically cohesive whole. Without this perfect synchronization, the result would be musically nonsensical or chaotic. Maintaining this "global musical understanding" across dozens of tracks is one of the biggest challenges in AI today.

Data Availability and Training Complexity:
Most AI models are trained on vast libraries of finished stereo songs (MP3s, WAVs). There is far less publicly available, clean multitrack data (stems) in the required quantity and diversity.
Training a model for native multitrack generation means teaching it to understand the complex relationships between tracks. This requires not only different data but also fundamentally new training methods that are extremely expensive and time-consuming.

Interactivity and Latency:
Features like "freezing" tracks (latent masking) demand near real-time interaction. The AI would need to be able to lock a part of its "brain" (the frozen tracks) and re-calculate the rest almost instantly without breaking the musical structure.
With current technology, the latency—the time between your command ("change the drums") and the audible result—would be far too high for a fluid creative workflow.

Conclusion: Implementing native multitrack generation is comparable to the leap from still photography to real-time, interactive 3D modeling. It's not just a matter of adding "more servers"; it requires breakthroughs in core AI research, data processing, and hardware efficiency.

[wintry pond]

in "Which versions sounds better"

Please let us mark if a generation is broken, and not just better

[clever harness]

You solution is kind of how it works today.

We don't stem split as a post process. It's generative.

[blissful sluice]

Is the stem process you describe as 'generative' technically a 'regenerative extraction' from a previously generated stereo mix? Or are the individual tracks (as with true native multitrack generation) recalculated completely in parallel and from scratch without a stereo mix template as an intermediate step? I'm interested in whether the stereo mix is ​​the indispensable data basis for the clean recalculation of the stems.

[clever harness]

I can't get into specifics, but it's generative in the same way songs are generative.

[blissful sluice]

Okay, well, that says a lot. It's no longer simple post-processing, but it's not quite the major revolution we're hoping for and what I was asking about.

It's a step in the right direction, but it perfectly explains why the results still aren't perfectly clean for me as a discerning listener.

It can only get better... I'm excited to see what's next.

thanks

[clever harness]

There's something coming soon to be excited about in this area I've been told.

[cosmic fractal]

what ML RLHF formulas are you using i have one that rewards diveristy instead of a single reward token : https://github.com/leeex1/Quillan-Ronin/blob/a79c7ca0e8ecd997176bfd0e424ecaad001bdcee/Formal Papers/A Deep Dive into the Council-Calibrated Reinforcement Learning (CCRL) Framework.pdf

[jagged garden]

SUNO TEAM:
PLEASE ANSWER YOUR TICKETS!!!!!
I HAVE BEEN UNABLE TO SUBSCRIBE FOR THREE MONTHS!!!!
AMERICAN. TWO SEPERATE AMERICAN BANKS ARE BEING DECLINE DESPITE HAVING MONEY IN THEM!

SOMEBODY HELP ME. PLEASE!!!

[clever harness]

Email [e-mail]

or #ticket-help

[jagged garden]

I tried that. You guys aren't answering.

Please please please help me give you money!

I sent screenshots and video. I don't know what else to do. Who do I have to beg?

Link is the problem. Or it's on your end. My banks are not declining the transaction(s).

You guys have been ignoring my tickets this entire time.

HELP!!!!

[wintry perch]

CAN WE FOR THE LOVE OF SANTA CLAUSE GET OUR ACCURATE SEARCH FEATURES BACK!

We had them for a while, where if I put the word "BACK" into the search it pulled up not just songs with "BACK" in the title but songs with it in the song. It's not doing this anymore.

It worked perfectly back in 3.8ish... 4.0 broke it and it has never come back properly.

I will die on this hill, the search feature needs to be better! There are people like me who work on a song, stop work on other things and then work on it again 3, 5, 10 months later, Now every version of that song, we need to listen to and compare, but now we can't find it cause 650 files are in the way? TABLE FLIP!

I know I delete song versions I hate, and have incompatible lines and things like that, but I know I don't get them all and miss some sometimes. This wouldn't be a problem if I didn't care if my workspace wasn't organized. But I do!
So now the problem is the search is so broken that if I type "BACK" and it brings up 2 out of the 300 versions only, and one song named Pony that doesn't have the word "BACK" in it at all(Looking at you trash can search, you are utterly useless) what good is a search feature like that?

WHY?! JUST WHY!

I begged for the search to be added to trash, I never gave up on that dream, and you implemented it, after me saying it in here pretty much every update. I'm not letting go of this dream. I will say it every new update.

PLEASE!!!!!! I AM PARKING ON THIS HILL AND NOT MOVING! @DEVS

[spiral bronze]

This is why I keep asking for better workspaces, because not only do we need a better search but we need a better way to organically sort tens of thousands of versions of songs that sometimes were worked on or returned to over months. It's only going to get worse as the handful of suckers like me who chose to buy Premier rack up more and more generations. I don't delete very many things, I just thumbs-down most of what doesn't work, but since there's no multi-select or anything it very quickly becomes impossible to change your mind about any organizational decision.

[kind grove]

It would be great if we could create workspaces within other workspaces. This would allow us to organize things in a structure like the following:

AmazingProject1
→ AmazingSong01
 → GodButNotWhat_I_Want
→ AmazingSong02
→ ...
→ AmazingSongNN

[torn hare]

Proposal – Persona and Release System

[prime sonnet]

How to delete an account?

[spiral bronze]

Presumably you mean your own account since it's probably outside of scope to suggest deleting somebody else's account.

[spiral bronze]

Subfolders are my #1 suggestion, alongside allowing multiple workspaces to be open at once side-by-side to make inspirations and mashups easier.

[sterile marlin]

🚨 Feature Request: Better Support for Traditional Chinese (Mandarin) Lyrics – Stop
Auto-Detecting as Cantonese!
Hi Suno team & community,
Problem:

• When inputting Traditional Chinese (繁體中文) lyrics, Suno almost always
  generates songs sung in Cantonese (粵語) instead of Mandarin (國語/普通話).
• Adding tags like [Mandarin], "國語", or "華語" in prompts rarely works – it still
  defaults to Cantonese pronunciation.
• This is a huge issue for creators in Taiwan, Hong Kong, and overseas Chinese
  communities who write in Traditional Chinese but want clear Mandarin vocals for
  worship songs, pop, or ancient-style music.
  Examples:
• Input: 繁體歌詞 like "我背著塵世的忙，推開一扇柴門"
• Output: Cantonese accent, not Mandarin.
• Workaround: Use Simplified Chinese (簡體) input, but that's inconvenient for
  Traditional users.
  Requested Fix:

1. Improve language detection: Detect Traditional Chinese → Default to Mandarin
   vocals unless specified "Cantonese".
2. Stronger prompt adherence: Honor [Mandarin vocals], "普通話人聲", etc.
3. Language selector in UI for lyrics (Traditional Mandarin / Cantonese / etc.).
   This would unlock way more Chinese creators! I've seen tons of similar complaints in
   Facebook groups and Reddit. Upvote if you agree!
   Thanks! 🎵

[weak sphinx]

Collab:
When you collab with someone, I believe you should be able to instead of only doing [Song Name, Ft. [Artist]] instead it should be similar to Bandlab

Sure you can do the title. But you can also have then as a featured artist next to your name so they too have credits to the track to:

Promote it on their profiles as a new song
When they reply to comments or comment on the track, they have [Featured Artist]

[arctic void]

I am having an issue linking to a private playlist. In the past, this worked just fine, but now the links just take me and those I send the links to to the main Suno page. I really need to share playlists I'm working on with others so they can provide feedback, so it would be great to get this fixed. Thanks! 🙂

[bold fulcrum]

Hi Suno, DOn't know if it's the good place but the Bot reviewing Hooks has a serious problem, or it has a 50's censoring system or don't work properly, I don't know, but I can't upload and share clips that's have no problem on youtube for exemple, and Youtube have already a strong censor system so whats the problem here with people dancing or enjoying summer ?
i post here the youtube link, I can't even post 20 seconds, it say it restricted content😩
https://youtu.be/Pd0a93hHnrI

And here the only Hook i managed to post :
https://suno.com/s/7GkW6lYDt4ASnCbV

[hollow locust]

Whoa, yeeeahhh... Mmmmm .... Yeah.

Suggestion: Train that out of the models intro on every song. If I want it, I'll put it there. I don't want what I didn't ask for.

[normal ember]

Need more vocal range and style. Constantly getting breathy billy eilish and pale Suno guy.

[clever harness]

#📚┃suno-school might help

[spark geyser]

Personally I would love if I were able to "deselect" instruments that would never be used in generations, a lot of same synths and sounds easily as if there is a small soundbank only available

[crisp ember]

Please do not translate the things shown as inspiration...
There are things that do not make much or any sense at all or are just "funny"...

Metallfalle
Zeichenabschnitt
experimentelles Gestein
Schreigesang
scharfe Schlingen
tiefes Messing
Felsstabil
Latino-Falle
Bruderland
verträumter Hall

[tough thunder]

100% this. The Korean translations are just as broken. and the ultimate irony is that if you actually use those nonsense translated words, the generative model has no idea what they mean😆 😆 Just leave them in English!

[oak fog]

Disable likes until the visitor has actually listened to the song. People are liking and leaving comments without listening to the song

[upbeat mantle]

That’s actually a solid idea it could genuinely encourage people to engage more with one another. though the problem whit that is would the like button only appear after fully listening to it or at the start.

[river tide]

not to mention they'd have to create a different system for the songs you as the creator like.

[spiral bronze]

Why not just make it so you only have to listen to about a quarter of the song to like it unless it's your song, and then like and dislike is free within workspaces?

[oak fog]

The like button should not be visible. Users should also not be able to leave a comment until they have listened to at least one minute of the song.

[clever harness]

If you had to guess, how much would you imagine this would impact your account?

In terms of a % of your likes and comments if that rule were implemented. I've never thought to ask someone what they the impact of imposing some kind of rule like this would be on their numbers.

[rare grove]

First, the bugs in the tracking system need to be fixed. It happens to me quite often that I listen to a song all the way through, but it doesn't register as a listen. On the other hand, I've frequently experienced the phenomenon that, especially with longer songs, the listen was counted three times (and I'm certain that no other two people listened to the song besides me, because then only my like was registered, and this also happened with songs by people with few followers). By the way, the follower system is also buggy. You lose followers without any action on your part or on the part of your followers (who then reported this to me), and about a quarter of your own followers don't appear in the list when they release a new song. The existing bugs should be fixed before implementing new features like the new UI, which will then cause new problems and frustration. (To be honest, I've been wondering whether I should even bother writing anything here anymore, since it seems pretty pointless.)😐

[clever harness]

Listens are abstracted away and not real time by design, this can make you think things aren't counted but it's all calculated.

[rare grove]

However, the proposal to only activate the Like button after a certain duration of listening to a song is then hardly feasible.

[clever harness]

Imagine a world where a like button appeared after 1 minute of listening, in this world, what should happen to users that defeat this protection?

[oak fog]

This isn't instagram you need to take the time to listen to the song before like and commenting

[clever harness]

Why is this important to you?

I don't disagree, I'm curious about your take

[oak fog]

because you can create a bot and like a whole bunch of songs

[clever harness]

But you can also create a bot to listen and then like and comment

[oak fog]

takes longer to trigger listen, likes are quick and spamming comments are easy

[clever harness]

Do you take advantage of blocking today when you suspect this is happening?

[oak fog]

Of course i would need SUNO metrics or back-end analytics to see, I am sure they have Google Tag manager to tirgger this as well

[clever harness]

If you saw far less engagement, but engagement you knew were based on this rule, how would you feel?

[oak fog]

I want the engagement to be more honest

[clever harness]

Me too

[oak fog]

Think about it, Suno has investors they need to show legit metrics with engagement and honest engagement

[clever harness]

It's an interesting idea.

[oak fog]

SUNO can also do some A+B testing on this as well

[spiral bronze]

I still don't know why, internally within a creator's workspace, the thumbs up-down system isn't a more granular 5-10 star rating system. It'd be more useful when looking for unpublished takes to know which takes I thought were better than others. Maybe even break it out into scoring across like, fidelty + other metrics in the case of bug reporting on tracks.
As soon as a song gets published, it can be replaced with the thumbs up-thumbs down system and listeners don't see the star ratings.

[oak fog]

Star Rating needs to be re-defined so the artist knows what to work on more for instance Vocals should have a rating lyrics I don't know its a maybe or could be adding more complexity as well

[spiral bronze]

I just know it gets irritating having to remember which out of 45 different thumbs-up versions of a track I haven't published yet were the meh but good melody ones, which had a cool guitar solo but the rest was boring, and which ones would be good but still need a remaster, or which ones were utterly borked but I'm keeping around to remix into something unrelated yet fun later. I don't actually care if it's a star system exactly, but organizing by title alone gets really annoying really fast.

[oak fog]

This would be more of a tagging system if it’s your own workspace and working on music

[spiral bronze]

Thus where the second part of my suggestion went into breaking out the star system into categories, so you can sort for crappy vocal gens, good vocal gens, etc.

[noble glacier]

In Realme9, Android 14, Realme UI 5.0, I can't create persona/voice. The save button is not showed so it can't be clicked.

It happened both in mobile and desktop view version. But I have no issues when I use older phone, Realme 3.

[bronze granite]

I’d like Suno to be able to automatically generate playlists from my public tracks, organized by genre, and then allow me to refine them using preset filters such as mood, language, and more.

[fierce tangle]

Hi Suno team, I’d like to give feedback about the current lyrics experiment.

My main issue is not just the UI design. What I really need is the previous Auto Lyrics mode and the workflow it provided.

Until recently, I could use Auto Lyrics in a much more direct and efficient way. The previous Auto mode allowed me to quickly generate lyrics and control the song creation process without extra steps. It was one of the most important features in my workflow.

With the new lyrics experiment, the interface has changed to the wand-based flow. I understand that this may be part of a test, but for my workflow, the new version does not replace the previous Auto Lyrics mode. It feels slower, less direct, and less useful for creating many songs quickly.

I use Suno for more than 10 hours a day, and Auto Lyrics mode was essential for me. This is not just a small design preference. Losing the previous Auto mode has seriously affected how I create songs.

As a Korean user, I also share Suno tips and useful information with other Korean users. Because of that, this change affects not only my personal workflow, but also the guidance I provide to other users.

I would really appreciate it if the Suno team could consider bringing back the previous Auto Lyrics mode, or at least giving users an option to choose between the previous Auto Lyrics workflow and the new experimental lyrics flow.

Possible options could be:

1. Bring back the previous Auto Lyrics mode.
2. Allow users to opt out of the lyrics experiment.
3. Add a toggle between the classic Auto Lyrics mode and the new wand-based lyrics flow.
4. Keep the previous Auto mode available in Advanced mode.
5. Allow heavy users or selected users to continue using the previous Auto Lyrics workflow.

Again, I’m not only asking for the old UI design back. I’m asking for the previous Auto Lyrics mode because it was extremely useful, fast, and important for my daily creation process.

Thank you for reading my feedback. I sincerely hope the team will consider this.

[distant sigil]

<@&1082792303976001547>

[rustic ivy]

I had a fun idea for randomly finding songs to listen to on suno.....its kinda like the magic 8 ball but instead its an image of a 8-track tape with a label that says Magic 8 Track. When you click the label a completely random published song on suno plays with the cover image appearing on the 8-track. It would be a fun addition to the webpage I think.

[hollow geode]

hi can someone help me? I cant login to my suno account on the web browser. While i am logged in on my phone. I use the same login on browser as my phone but as soon as i log in i get a new account. I just want to log in into my paid account..

[delicate cedar]

Are you able to successfully login while in an incognito tab?

[hollow geode]

let me try

nope

[delicate cedar]

What browser do you use?

[hollow geode]

firefox

the same happens if i use chrome..

[delicate cedar]

All I can suggest at this point is #ticket-help sorry.

[random egret]

I'd like to suggest a new feature: extending songs BUT backwards. Right now we can only extend tracks forward. But what if you want to try something unusual and create a completely new start or intro for an existing part? It would open up a lot of creative possibilities and i'd love to see this feature added! Don't know if others will like it, but it would be awesome

[delicate cedar]

Studio kind of does this.

You have to drag the track back and then prompt an empty space that you lasso.

[random egret]

Oh! You actually right... thx

[kind grove]

But for the Editor (Pro Plan) will it be nice too.

[coarse wyvern]

So I wasn't sure were to put this but it is a suggestion for Suno i would love to see integrated for when making music I know you can use prompts to have something change in your music but what would be nice say if i wanted to do a trio if you can do something like [verse 3 | Use Persona] and then on that selected verse only your persona will be used for lyrics cause I am finding it extremely difficult to make a trio song what I'm getting is 1 female voice and 2 male voices that are exactly the same

[solar sedge]

a "log out of all devices" option

[pulsar mulch]

Would have loved a more expensive generation where SUNO built the stems as it generated the song instead of mixing them all up, making the function more or less useless now. I generate folk music and the studio and the stems are just random sounds put on different tracks making it un-usable to clean up since other instruments ends up in the vocal stem etc.

Would also have loved a more conversational style of generation where I could tell it that a certain song felt better and then it build from that instead. Now its a new try everytime you press generate.

[marsh lichen]

Listen for Credits feature was available for a long time with 15 seconds per track. Several days ago, it switched to 5 seconds per track. Now it's suddenly 30 seconds. Nope, guys, I will hardly use it now. The best time would be 10 seconds. Most times it's enough.

[patent gulch]

Can't log into my Suno acct via code. Tried incognito, different browsers, nothing, code no longer comes to phone

[shadow inlet]

I would really like to be able to force a minimum length for the "extend" feature. I often have to create multiple versions to have something in which lyrics are not said too fast.

[rustic ivy]

I feel that you should be able to download someone elses cover of your song.....after all it is a trade off and that would only be fair

[teal shore]

make a loading bar for videos like what has been done with wav files, so we know how long it's going to take

[teal shore]

also, it would be good if, when you make a new section in the Editor, you can change the tempo of that section instead of it changing the tempo of the whole song. that would be great. so like, a 130bpm song dropping into a 145bpm bassline

[teal shore]

also, if we could change shape of fade in/fade out, the sine wave default is ok, but there are different variations in daws that allow some songs to fade out better than that (a lot better)

[gusty granite]

I would love to see the ability to use more than one Voice to set up duets or even trios

[marsh lichen]

Ideally, we should be able to choose any number of Voices in one song. It would be really useful for those who create rock operas or something like that.

[crisp ember]

I just accidentally reported a wrong song as "I just don't like it". I'd like to have an option where I can see, what I reported for what song, and furthermore the ability to change or cancel/delete a report (Of course not days later...).
And it would be nice, if it would not send a report immediately, maybe should be check boxes.

[bleak ruin]

You currently offer Simple and Advanced modes, but there’s still something missing a true Professional mode.

What we need is an option with no safety nets. A mode where the AI doesn’t try to decide if something sounds “good” or not, and doesn’t interfere with the creative direction.

This mode should give full control over how the music is structured and performed. No extra fluff, no forced interpretation just a large lyrics/instruction window where we can define every part of the song ourselves, directly and precisely.

Right now, even Advanced mode still feels guided. A Professional mode should feel unrestricted, built for users who know exactly what they want and don’t need the AI to “help” shape it.

I’m not asking to remove safety limits. I’m asking for a mode where the AI follows user input more strictly and interferes less with structure,

[gusty granite]

Yes that would be great!

[hot zenith]

Would be lovely if any of the features that come with pro actually worked.

Cover with voice

Add instruments

And

Add voice

All do absolutely nothing.

Why am I paying for these "features"? If they don't do anything.

[indigo rivet]

I'm using other AI music tools that have a vast array of voices you can choose from for each musical genre. A lack of choice with Suno is a serious disadvantage. Please address this. Thanks. 😥

[fervent pecan]

here is place for ask or say something right?

[sour shore]

not that anyone's gonna read this cuz the community and dev team here don't know how to communicate but:
v5.5 STILL (after several months) doesn't generate anything over 1 minute while using cover

[clever harness]

Do you have an example link?

[sour shore]

to a generated song?

[sour shore]

I seemed to have thrown them in the trash along my wasted credits

[amber violet]

After generate stems, there should be a feature that regenerates the stems clean using AI to enhance their quality individually. One with the same level of echo and reverb, and one without. This would be great to enable us to mix the channels better ourselves and have better sounding quality.

[karmic sigil]

Unless you went and perma-deleted from the trash bin - you can recover them and link them...

When making voices - let us crop out/delete sections before finalizing - or give us multiple selection contexts so we can pick different parts of the song's vocals and skip over stuff in the middle if we want or think it's bad.

Essentially, let us only pick the literal best parts of the voice/song - not have to keep one contiguous piece that might have bad sections just to get all the good ones.

[tepid laurel]

Expose the Duration parameter.

I should be able to control the duration. Which controls how fast it sings the lyrics. That way I can make it take it time. Or sing fast.

[loud kite]

Once you have a voice chosen by the model for a song it is difficult or almost impossible to get rid of it and change it with another oner, even using personas/voice and making covers. That is a big problem and that seems to aggravate with each new model. I don't how other solve this problem.
I just tried to replace a singer using cover and 4 different personas and the initial voice is still there.

[heady wind]

Hoping this is the right place to raise some suggestions for product improvement? In the ReMaster feature, you have Subtle, Normal and High as options - can you add in a Clean option? Often times a track is nearly perfect but there are places within the track where the effects are just too much and it's impossible to correct this in an external DAW. The Clean function would review the track and remove any overzealous effects. This would really help in getting decent tracks downloaded.

Suggestion 2 - Standardized download stems structure. Currently when you get the the stems downloaded, it's really piecemeal and sometimes you don't have enough separation to allow professional external mixing. Ensuring that all stem extractions follow the below format would alleviate many of these problems and massively improve Suno as a professional music making tool;
01 - Kick Drum
02 - Snare Drum
03 - Hi Hats
04 - Cymbals
05 - Additional Percussion
06 - Bass
07 - Keys 01
08 - Keys 02
09 - Guitar 01
10 - Guitar 02
11 - Lead Vox
12 - Backing Vox

[spiral bronze]

That's kind of what remaster is supposed to do already, it's just not very consistent at it. I wish we could give remaster a prompt telling it exactly what we want it to do.

[heady wind]

Indeed - a 'Remix as is but' function would do. Sick and tired of doing 30 subtle remasters because the lead guitar sounds too boxy on the Middle 8 - 🤣

[spiral bronze]

Yeah, when v5.5 initially dropped every single thing I genned was a complete wall of static in the second half and I was blasting through 300 credits on a high remaster per track just praying to RNGesus for something slightly less hazy. I find subtle remaster basically useless since it doesn't clean enough most of the time. I know somebody suggested before if they'd let us remaster stems after splitting them out. I shudder to think of the credit cost but then again it'd probably end up being cheaper compared to what we're already doing.

As for your stem suggestion, I'd honestly say they should let us tell Suno how we want it to use its 12 tracks. Because I may want that exact layout, but if I'm doing a different genre I might not care about the drums in that much detail as much as I just need the snare by itself and maybe I have some synths and weird effects as well as multiple lead singers or something. But right now, the bleed between tracks is really sketchy and honestly I think we just have to wait until we can get a model that's clever enough to work in Studio.

At the moment, Suno can't decide if it wants to follow your stem prompts or if it needs to generate full songs even when told to just regen a stem. I can't imagine that bodes well for any "clean stem" function.

[heady wind]

Stems - you may be right, I just thought we might get one step closer a little sooner, rather than waiting ages for a complete overhaul. 😀 Remasters - Do you think that the remasters get better the more you do on a song, or is it really just random? Also, have you noticed any difference in overall track quality based on the time you create it? Eg - a Wednesday morning track will usually sound a lot better than one produced on a Saturday afternoon - probably owing to server loads?

[spiral bronze]

I don't think Suno "remembers" how its already remastered a track, but if you remaster a remaster, I'm not sure. I imagine that it could help if you get like, one issue solved in the first remaster so you decide to remaster that one to iron out the other one. I've never needed to do that thankfully.

I know there's a little bit of voodoo in this community as far as sometimes claiming that the model "changes" how it reacts to prompts, changes quality, etc.. I think a lot of it is a combination of placebo and changes to the way Suno processes prompts. Most of these web AI services don't actually send your raw prompt straight into the model, they fluff it up to try and give you a better quality result or to add their own special sauce to it. It's basically just like them clicking the "enhance prompt" button for you, and I believe that they change that algorithm and its tweaking far more often than they change anything else.
It's possible that they change the model even within "v5" and "v5.5," but it wouldn't be like raising or lowering the quality of it, because that'd be awful for business if they ever got caught doing it. Wouldn't make any sense from their perspective anyway, that just increases training time to make multiple quality levels, and 90% of the cost of running an AI company is in training.
What's far more likely is that they have different flavors of each model version, and they swap them out occasionally to test what sounds best. See, AI models have to train for X amount of time, but after that you don't have one version, you have maybe eight versions that all ended training at different times. It's like aging spirits in a barrel. They're all fraternal twins, very similar but a little different.

[heady wind]

Interesting. I do know that about 4 or 5 weeks ago, after 3 months solid results (or rather results of consistent quality based on input), for about a 7 day period I couldn't get a single decent track out of the systems. I really thought they had fubarred the algorithm with some unannounced update, but after that period it came back and now it's pretty consistent again, though I still think that tracks produced at low traffic periods seem to turn out better than ones done at the weekend. Could be confirmation bias though - I'm a musician not a scientist.

[spiral bronze]

I've experienced the same thing. I'm not a musician at all, but I am an audiophile and I have very expensive equipment for the listening side of things, and I've always harped on about the need for good fidelity in music and sound design. I dabbled in sound design as a hobby before AI. Suno definitely has these moments where it "works" and then other times it has these weird out-of-phase artifacts, I've had percussion get these ghosting transients, and lately I keep having lead instruments sound underwater by the end of the track. Not to mention the off-and-on volume creep 5.5 does now.

One thing that's really hard to grasp with generative AI is the way training works and how that gets communicated to the public. See, I kind of alluded to it before, but when you're training an AI, it's not like you're starting at 0% and when it gets to 100% you're done. You basically get your data, get your structure and your fancy bits, and when you train you kind of just sample it every so often and see how it's going. And you wait until it either seems like it's improving or you stop and start over with different settings. Well... When you get to epoch 500, maybe that sounds great. And maybe epoch 800 sounds better. Maybe epoch 1,500 sounds amazing. But then epoch 2,000 sounds terrible. And epoch 2,500 sounds worse. And that's overtraining, it means the model is overlearning lessons you weren't trying to teach it. You prompt for "Classical" and now it HAS to insert a violin. You prompt for "soft rock" but it learned that rock is always about screaming. It just doesn't know any better.

But beyond that, when you get into the subtle aspects of it, you might have epochs 1,400 and 1,500 and both of them sound fantastic. Which is better though? Hard to say. If you test only on rock music, 1,500 might sound better of the two. But maybe 1,500 makes worse reggae and it makes really tinny awful reverbs, which 1,400 is much better at. They're all the "same" model, they had the same training data, but one overlearned some lessons and underlearned others.
No way to know every single nuance without putting it out there and seeing what it does. Then you collect the data, regroup, figure out some new tricks, and start working on the next model.

[pale imp]

How do you get Spoken intros in 5.5?? Every time even prompting it it still gives me drums or guitar shreds after every finished sentence...

[spiral bronze]

The way I do it is with square brackets []. It uses the first line for internal prompts and then speaks every bracketed line after that (even without you saying "Spoken" but I do it for consistency).

Example:

[Spoken:]
[This is a spoken intro]
[You can write as many lines as you want like this]
[It will speak all of them]

And then it sings this part

[spiral bronze]

As long as you keep the lines together in one block it shouldn't try to put stuff in between them, but you might have to try a few times or shift your other lyrics-instructions around.

[heady wind]

Very interesting. And if you think AI music generation is frustrating - have you tried AI music video generation? I swear that will break even the hardiest of psyches. 🤣

[spiral bronze]

Coincidentally I just started doing local video generation today, and if I didn't already do local image stuff I'd be up a creak. Writing prompts is hard, and I'm on easy mode since I'm just trying to animate images I already made not describing a whole scene for it.

[heady wind]

Well I've been through most of the currently available AI video makers and I settled on Neural Frames. It's not cheap and I usually have to do 3 videos to get enough usable footage for 1 video, but when it works, it works really well. But sometimes it just gets massively off track.

[spiral bronze]

Yeah, it's always a balancing act. It's like prompting in Suno, sometimes you just have to let it do what it wants to do and treat it as a discovery machine rather than a "translate my imagination into reality 1:1"-miracle maker.

[heady wind]

100%. When I first started with Suno I was uploading prerecorded full production 32 track precision engineered masterpieces and writing 1000 character style prompts - now its usualy just me and an acoustic guitar or piano and a 50 char prompt and the results are usually much better. I think I was over-egging the cake & now I like to give Suno the room to breathe and stretch it's artistic legs.

[spiral bronze]

That's the best way to do it. There's a time and a place for the ambitious products, and it'll probably be a lot easier in a few years, but it's always best to play the tools' strengths. Or at least its quirks. And Suno sure has some quirks...

... thus why I largely put downtempo and shoegaze influences into most of my tracks. If I'm going to get crappy audio quality, I might as well make that a selling point!

[haughty topaz]

Add login method (email/password) I cannot log in through Google Workspace.

[pale imp]

That's what I've been doing.

[loud kite]

SUGGESTION FOR DESKTOP VERSION
It would be good to have the line highligted in the lyrics where the singer/player is just singing...
And also to jump in the song at the exact line that you click on in the lyrics window (while you are in the create window).
There is an AI who has both features implemented.. and this makes a difference.

[surreal glacier]

only way to get reseults is - MJ mood boards / character sheet > seedance 2 + upload vocal stems to sync = But then your vid gets blocked by SUNO moderators like this one https://suno.com/s/SC29hd96AfwHkTRx - they have blocked 15 vids all vanilla, choreography 🤯

[blissful sluice]

🎙️ Feature Proposal: Persistent Voice Profiles with Full Parameter Control

TL;DR – Suno Needs True Voice Profiles, Not Just Voice Labels

The Core Problem
Currently, a "Voice" in Suno is just a weak label. As soon as genre, mood, or instruments are entered in the prompt, the model completely overrides the voice. Result:

Voices sound different from track to track
It's nearly impossible to consistently develop an established character
Voice descriptions lose out to the rest of the prompt

The Vision
A Voice Profile must function as a strong pre-compiled prior – a fixed anchor that takes precedence over all other prompt signals (similar to Eleven Labs). If I write "Voice: Harald," Harald should remain Harald.

The proposed solution – Three levels

Level 1 – Identity (foundation)

voice_name, gender, age_group, voice_type, pitch_range, primary_language, accent_region + strength, etc.

Level 2 – Acoustic & Tonal Profile (the actual fingerprint)
Timbre (brightness, breathiness, raspiness, nasality…), formants, vibrato (rate, depth, onset), dynamics, pitch stability, etc.

Level 3 – Behavioral Parameters (the underestimated gold)
Articulation, melodic movement, technique toggles (melisma, falsetto, growl, belt…), emotional defaults, context override resistance (0–100) ← Key parameters

Additionally: Phonetic Language Profiles (Part 2)
Dedicated system for language variants (e.g., zh-TW-Mandarin vs. Cantonese, Arabic dialects, etc.) + Community Reference Packs – structured Audio uploads from native speakers who teach Suno true phonetics.
Why this must happen as a saved profile and not in the prompt:

Prompts are context → everything competes.

A saved profile is a fixed prior that is loaded before generation. This fundamentally solves the consistency problem.

[worn ivy]

I have exactly the same issue as you. Even tried on a MacBook but keeps opening a new account and not my pro account. Really annoying

[clear sun]

Suggestion: It would generally be a good idea to expand the payment options.
I’ve been waiting ages for the option to pay via PayPal.
You’re missing out on so much revenue because you haven’t added this option yet.

[hollow geode]

Did you get any help yet? I didnt

[loud kite]

EXTEND: Is there any logical reason for the extend length defaulting to half of the uploaded songs every time?
Why not at the end?

[ancient matrix]

any updates on official suno API support release timelines yet?

[naive wagon]

Has suno every thought about adding an option to choose generation length?

[deft vapor]

Suggestion: Different Audio Influence Sliders. Theres only one but yet I've noticed you can give a song an insporation list, a voice file (for using your own voice) and you can also upload or give it an Audio file for whatever other reasons you may need to.

Thing is however, we only get one additional slider for "Audio Influence and it combines all of it into that one Slider. I feel we could have more fine-tune adjustments if we split that slider for each audio type. One for the Voice, one for if you're doing inspo or remix or whathaveyou, and one for when you upload an audio type.

All I'm saying is if you're going to give us 3 MAIN audio usage points, we should have at least 3 seperate bars to adjust the usage of thoes 3 main points, not one for all

[deft vapor]

That would be a nice feature tbh, but usually length depends on the genera, emotion/tone the song is portraying (sad and emotional ones usually being longer I've noticed), lyrical structure and length of all your lyrics, and my "fav" the occasion that Suno decides to replay the song back at the start or at some point inbetween... so looping, which i've noticed can mostly be stopped by simply putting [End] at the end of your song/lyrics.

[spiral bronze]

<@&1082792303976001547>

[worn ivy]

Nothing. Sent messages to support and never got a reply.

[dense otter]

The fact that your AI has double-standards for Hooks, allowing some idiot with a bird chest to be shirtless, then moderate an image of a man wearing a traditional vest dancing is fcking pathetic, useless, obtuse, and just plain worthless.

[fleet echo]

Creator Ownership + Download Access Layer (Workflow Gap)
Quick product observation from using Suno for full track creation workflows.
Issue
Suno currently works extremely well for:
generation
sharing/streaming
discovery
But for creators treating outputs as finished assets, there’s a missing native path for:
downloads / export
external distribution workflows
catalog building (long-term creator use)
This creates friction when moving from “creation” → “real-world use”.
💡 Concept: Optional Download + Supporter Layer
Keep streaming/share as default, but add an optional monetization layer:
Paid users can download tracks from supported creators
Creators stay fully independent (no ownership change)
Fans can directly support specific creators inside Suno
Streaming remains discovery-first
🧱 Simple model (conceptual)
Free → stream/share only
Paid → download access for supported creators
Optional creator controls:
enable/disable downloads per track
supporter-only releases
export quality options (MP3/WAV/stems optional)
📈 Why this matters
improves creator retention (less off-platform exporting)
increases subscription value without changing core UX
enables direct fan → creator support loop
preserves Suno as discovery + creation layer
supports “finished track” workflows inside ecosystem
🧭 Core principle
Keep Suno as creation + discovery
Add optional ownership/export layer for serious creators
Just sharing as a workflow gap observation — not a request to change direction.

[trim quail]

got a suno suggestion!

male vocals and female vocals singing together

option-basically

[stray spindle]

To the development team, is there a chance that SUNO will release a CLI tool in the future?

[teal shore]

make it so that [End] ends the song where you put it, and when not many digits have been used. i can make 4min tracks without a sweat on earlier versions but it's seemingly impossible to do on 5.5 😮‍💨 it's like playing a gambling machine with a low win rate

[golden acorn]

Hey SUNO AI team. Wondering whether you would consider implementing FLAC format for the Download feature. It has decent compression and lossless quality. And from my experience, it shouldn't be hard to add to a web app. Did it in the past for a small project: https://github.com/EnthusiastGuy/S.E.A.M-Audio-Audition

[hushed lotus]

Hey fix your site

Quit suppressing the people's reach, put actual good songs on showcase.bc your taste are terrible

Fix your loops ,I have to contact the ones laying around in Massachusetts, sell.it back to China when it was actually fun

Fix the loops.
Fix the vocal drift.
Stop burying actual creative tracks while showcase gets flooded with generic copy-paste songs.

People are burning credits fighting the model instead of creating.

[hot violet]

Fix the SEARCH engine

[teal shore]

update listen and rank so that it looks like this, so that i can actually see how many credits are pending when i'm listening/earning:

you got all that free space on the screen, may as well use it up, pretty sure that's a basic thing taught to people who learn web design

[rare grove]

Hello, my friend—I agree with you 100%. Unfortunately, posting anything here is pointless. I’ve done it repeatedly, but evidently, they have no intention of changing things; instead, they just keep making them "worse in the name of improvement." We were told that this is all part of a grand, brilliant plan—one that cannot be revealed just yet—and that in due time, it will turn out to be magnificent. I, for my part, have taken the logical step and cancelled my Premium subscription (which, unfortunately, still runs until November). Aside from protest songs, I’m no longer publishing anything at all. I know that many other people are doing the exact same thing—or are simply seeking out alternative platforms, such as YouTube.

[pallid dirge]

Hey Suno you know what il really like is allowing midi imports

[cinder jacinth]

I want to bring up something that honestly crosses a line for me with the new “turn your text messages into songs” push.

I already handled the personal side of this directly with the person involved, so this is not a callout toward them. The bigger issue is the feature itself and what behavior it normalizes.

Conversations between people carry implied trust and context. Turning normal private interactions into viral content templates encourages people to treat conversations like raw material instead of human exchanges.

I recently had my own words from a real conversation turned into one of these songs. No names were exposed, but it still felt wrong hearing a personal interaction transformed into content without being asked first.

That’s the part I think Suno seriously needs to think about.

Not every interaction should become engagement fuel.

A trend being viral does not automatically make it healthy. Features shape behavior, and this one pushes people toward airing out conversations for views, novelty, and reactions without really thinking about consent or boundaries first.

What also concerns me is that this is being rolled out very casually and even free-access promoted, which makes it feel less like a carefully communicated feature and more like live behavioral testing for a larger platform push.

I’ve criticized UI instability and visibility changes before, but this hits differently because this involves interpersonal trust, not just workflow.

Creators and users are still people first. Conversations should not automatically become content pipelines.

[cinder jacinth]

Continuing this thought, the thing that worries me even more is that this feature is currently being pushed for free and heavily framed around viral behavior. That makes it feel less like a harmless novelty and more like live behavioral testing for future platform integration.

The easier and more frictionless this becomes, the more people are going to start treating real conversations like content material because they think it is their next viral moment. That directly changes how people interact with each other on the platform. Instead of conversations staying conversations, they start becoming potential engagement bait.

What makes this feel dirty to me is that users are effectively training and normalizing this behavior in real time while the platform gathers usage patterns, reactions, trends, and interaction data around it, but without openly framing it as an experimental feature rollout or behavioral test. It just gets presented as fun and viral.

I am not saying nobody should ever use the feature, I am saying there needs to be a serious conversation about consent and interpersonal trust before this behavior becomes normalized across the platform.

[autumn rose]

Make a Discord Activity of Suno. Where people can listen to tracks. I would like to code that as a Discord bot developer tbh

[proud cape]

Suggestion:
Points refund for undesired results.
95% of generated results are not acceptable.
Without the refund, system is rigged to scam you with bad results and there is nothing I can do to prove it one way or another.
But we already know that's exactly what's happening.

[delicate cedar]

You're paying for the output, not the results.

If you're renting a kitchen, you're paying for access to the tools of the kitchen, the costs of running the oven and or stove and ingredients. Those cost to use.

Whether or not your cake or other recipe comes out is largely up to the skill of the chef. And personal preference. But the costs are the same regardless of whether or not the spinach souffle failed to rise or not. The costs are still due because resources are used, no matter whether or not you like what is made.

[proud cape]

I understand your reasoning.
How ever my argument is that bad results are given on purpose to milk your points.

[delicate cedar]

I'm not fond of conspiracy theories. Nor are most of the mods here on the discord.

[rustic ivy]

My prompts are written to where I get the desired results in 4-6 songs. It takes more if I try to go too crazy with it. Good scripting and preplanning is key.

Tho I wish suno ai would give us an inventory of their tools AKA encyclopedia of prompt tools and scripts that their system recognizes......heck a seed feature so that you can regen that track and micro edit like they allow with photo gen software

[clever harness]

It doesn't exactly work that way, which is why there's no prompt list we can provide people. If you rethink the prompt as just a math problem, and the song is the answer. There are an infinite number of things to put in the formula and an infinite number of answers.

Just like you can't provide every number in every equation and every answer to every equation.

[rustic ivy]

I kinda get that but surely the ai has an instrument list, genre list, with attached instruments to that genre....know what I mean?

[clever harness]

Think of it like massive math problem, and users agree on questions (prompts) and the answers (songs) via their actions when they get the answer. Those answers might make sense to the majority, but not make sense to the individual (you think the answer to rock is A, but everyone else thinks it's B). I could tell you "type rock," but I can't tell you it's the answer is what you think of as rock. You might also type Lynxsy as your question, and the answer might be rock.

[rustic ivy]

I understand, I really do 🙂 For me its like this....I have no musical education due to the fact that I am quite literally almost completely deaf. I am on the spectrum, and Im trying to use suno as a canvas....sound and vibration. Im guessing at sound colors, and guessing at command brushes. Its so frustrating that I cant make what I need cause I dont know what to use, or even capable of. Its like an art program like photoshop.....it has unlimited artistic potential, but all the buttons and tools are blank and you just have to guess. What command or coding to put where, to do what, will it even be recognized by the ai etc. But yeah, I get ya. I try to learn from folk, but many once they figure out some secret from the black box, they keep it to themselves. I will keep marching on, thank you for yer time and thank you to the developers for giving me the opportunity to let the music outta my head 🙂

[clever harness]

What I'd try in your position is look up reviews for some of your favorite songs / albums, and then use that as the prompt (probably will need to remove references to the album itself to get around filters.

[clever harness]

There's probably a way to visualize this as a labs project that would help someone with no musical education use language they do understand to find sounds they'd find pleasing even if they didn't understand the prompt... 🤔

[west elk]

I would like to be able to rearrange the songs in playlists.

I would like to be able to manage which playlists a song appears in, easily. On the web interface, if I decide to unpublish a song, there is just no way to figure out what playlists it is in. I have to open up my phone, search my library for the song (mobile search sucks, btw, apparently doing an "OR" search for all keywords instead of looking the title as provided), identify the song by comparing the album art to the one on my PC, opening the song, then going to add to playlist and from there, deselecting the song from everything.

That's way too fckin awkward. Please implement better controls to manage our music.

[west elk]

This button does not work. I have no idea how it's actually sorting the songs, but it not reverse-chronological.

Web interface. PC. Win 11. Please ping me if you want more information.

[merry flicker]

I want to link my Google account and phone number. How do I do that?

[dark leaf]

I gave this suggestion over 6 months ago so here it is again: I want a way to show where I want men and women to sing. For I am trying to make a song with both of them singing in different areas. One way would be to let the user highlight the areas where males sing and then highlight in a different color where females sing. The colors, for example, are blues for males and reds for females; when they are to be singing at the same time, the mixed color is purple to show that both males and females are to be singing at that time. Please make this possible! It's annoying losing credits due to trying to make the vocals sing where it is a male singing versus a female singing.

[west elk]

I've seen a lot of people use tags in the lyrics like [verse, male vocal] or [chorus, female B vocal]

[heavy trench]

Hey Devs,

When it comes to future versions, I would like to have one, but you can have the AI sing with multiple parts. I say if you wanted to do a Bulgarian harmony with like four or five voices, I did something so much that but I can only get like two voices out of it. And I’m not looking for like choir voices or anything. I’m talking about like multi layered voices.

[heavy trench]

Does it work?

[west elk]

Seems to. Ymmv.

[west elk]

Have you tried specifying "multilayered vocals" for the section?

This is why Suno going to the "narrative method" for music style was a bad idea. I still use the style-tag method to specify what I want and they work very well.

But people don't know to do that anymore because "you can just talk to it." Suno did the users a disservice with that shift.

[teal shore]

ask them to put some kinda info page about all the extra features on the site somewhere, explaining what "cover, remix, resample, mashup" etc actually do, because to me, all they do is waste my credits because i dont know what i'm doing with them.

[dawn olive]

Please try and better understand sub genres.
No matter what kind of prompt I input for post punk (late 70s to early 80s) British electronic Berlin school sound, I get that homogenised modern electronic sound.
Sonauto understands sub genres VERY well. Please consider these particular genres as they are distinct from the more US centric synthwave

[spiral bronze]

Cover = Remix.
Remix = Keep your song more or less the same structure depending on slider strength, but adapt it to a new prompt or change the way it's sung or other such things. If instrumental it will tend to vary playi g style more.
Inspiration = Combine chunks of your inspired songs together into a new piece, with no regard for structure.
Mashup = Keep the structure of two songs intact but combine them together.
Sample = A more controlled version of Inspiration but only for one track not four.

[spiral bronze]

You're going to get blapped for mentioning a competitor, but if you want a specific subgenre sound sometimes using a different term for it works. It might be confusing genres with similar names, like mentioning Berlin might throw it off. Describing the sound in terms of what that subgenre has might be your best bet.

[dawn olive]

Many thanks. I just don’t think the Suno model has much of that “old weird stuff”
In it that therefore generates the kind of sound I’m looking for. Hopefully v6 which will be off the licensed content will have more “indie vibes” in there but we’ll see. Currently I find suno too homogeneous.
As for competitors… competition is healthy! 😉👍

[karmic sigil]

Since we can't see "liked playlists" as a tab anymore to find them again - let us "Save" or "Bookmark" playlists in addition to like so we can find them again.

[marble arch]

Still add random play of playlist is still one of my top suggestions / requests both for the webapp and IOS.

[heavy trench]

I have, but that was when version five was out. And I just kind of like completely ignored it.

[keen marsh]

I think a megamix feature would be awesome—you could pick up to 5 songs from your library, and Suno would create a megamix out of them, with smooth transitions, matching BPM, and so on.

[mental hornet]

Suno, my workspace has too many song variations. I kind of want to just delete all the unpublished songs because listening to and categorizing all of them is too much work for my ears. 😭 Yeah, I’m just whining here. I wish there were a way to add tags to songs the first time I listen to them, so I could easily find what I need later.

Examples when listening for the first time:

“This could be used as a sample” → tagged [Sample]
“These vocals are interesting” → tagged [Vocal]
“This instrumental part is good and could be used for Milo” → tagged [Instrumental]

Then in search, we could search using existing tags. And it’d be even better if we could freely create our own tag names.

It would save a lot of users like me so much time, I think. 👻❤

[final drift]

Not sure if anyone has asked for a way to combine voices and create bands, if so *Bump! If it was possible to use vocie names in prompting for lyrics and the context could be realized with multiple personas I think the option for creating bands from mixed voices should be possible (just my thoughts there) Also I have found that if I select a voice and use the sound prompt for it (older pre-voice) then change to another voice I can get duets with both of those voices as a pre-workaround for something that to me sounds like a serious community grower...

[drowsy cosmos]

use the title of the song / create workspaces with those tags / easy.

[haughty nova]

Suno, can you please release an official API?

[mental hornet]

It’s doable, but it’s not enough. Cutting off the workflow just to move songs into another workspace or rename titles gets tiring fast. After 3000+ variations, can you even remember what your notes meant anymore?

Tags are better and save much more time.

Technology evolves so humans can do less repetitive work, and tags aren’t exactly a new concept. 🤘

[drowsy cosmos]

@mental hornet divergence/convergence is not that hard

[mental hornet]

Okay, got it. Your thoughts are yours, and my experience is mine.

This wasn’t meant to be a debate. I like tags because they’re fast and save time during the creative process. If you don’t, that’s fine too.

I’m just sharing my idea here. Whether it happens or not is up to the Suno team now, and that’s outside of my control.

So continuing this discussion isn’t really worth more of my time.

Bye~

[drowsy cosmos]

My "thoughts" are based on 25+ years of working in media entertainment industries (Netflix/Mtv/HBO), handling thousands of files per week. The solution is not a tagging/naming convention system but a workflow. I'm trying to give you a shift of vision and mindset so you can try it. What you want seems to be a perfect solution; however, you need to make a test and see if actually tagging saves you time compared to naming convention. Also, you need to slow down the process if you are making an assessment regarding the file intention. If you end up with 3K of possibilities, there is a flaw and no tag will help. I'm sharing that, a POV not a discussion.

[long robin]

Would love a slider for mashup that lets us tweak which % of each song to weight, I've noticed with some it will only ever sound like 1 of them so being able to tweak it so the other can come through some more would be neat.

[spiral bronze]

This is just a natural extension of the shift that happened ten years ago. Remember, they weren't always called "DMs", they used to be called "PMs."

[cinder jacinth]

Yeah but I think you're also missing the point of the topic here just hypothetically if this happened to you how would you feel

For me it wasn't hypothetical

[spiral bronze]

No I'm actually not missing the point. I agree with the message I replied to.

I'm saying that Suno is following a trend, a dangerous trend. They didn't set that trend, though, it's everywhere.

[delicate cedar]

Did you repost this?

[cinder jacinth]

They're not following trends they're setting them

[delicate cedar]

@cinder jacinth reposting the same thing is spammy and subject to removal. You posted this on 5/10/2026 and got multiple reacts (including my own) in agreement with you. No need to post it again.

[spiral bronze]

Hard disagree. I haven't seen Suno do anything that isn't trying to carve out their own slice of an existing market. It's not bad to follow trends, it is what it is.

[cinder jacinth]

I already made my point and I stand by it. I repeated it because I genuinely think this trend is crossing privacy and trust boundaries people are brushing past too easily.

If the repost itself is the issue then fine, I heard you. But I’m not backing away from the point I was making.

[delicate cedar]

I'm not a fan of the trend either. Despite enjoying some of the social media comments that have been turned into songs, it's not my favorite thing. This whole phone text thing is even less of a favorite thing of mine for precisely the reasons you said before. The privacy implications.

But yeah, any text that gets posted multiple times is considered spam. And spam gets removed when it's seen.

[cinder jacinth]

At least you understand the actual point I was making, because too many people are treating this like harmless internet fun without thinking about where this kind of normalization can lead.

My reason for repeating it was simple, one post disappears into the feed and people move on in five minutes. Serious concerns get buried under trend momentum constantly.

I get what you’re saying about the reposts, but I’m still standing by the privacy concern completely.

[delicate cedar]

I'm not disputing for a second the privacy concerns. I think they're real ones. (Then again no one consulted me on the matter before announcing this. Which makes sense since I'm only a Discord Moderator.)

The social media ones I mentioned earlier are less of an issue because they are posted publically on the web for all to see. A text message or DM has the illusion of privacy attached. It doesn't matter if it's an illusion. People still have the expectation of privacy in their private texts and DMs.

[cinder jacinth]

And that expectation of privacy is exactly the point I’ve been trying to hammer home this whole time.

Once platforms normalize turning private conversations into entertainment content, people slowly stop treating those spaces with the weight they’re supposed to have. That’s the part that bothers me the most.

I’m not blaming you personally for the feature existing, but I do appreciate that you’re at least acknowledging the concern instead of pretending it’s harmless.

[vivid dagger]

When can we expect to record in stereo in studio? A stereo track LR.
Are you working on fixing the delayed recording of a track?
The playback of a recorded track in Studio always starts late.

[wispy gale]

Can we have filter option in Creation Studio to HIDE Uploads?
So I don't have to cycle through my uploads to find my other songs whenever I've done a custom model?

[quick sequoia]

"Earn Credits" feedback. I have a complaint, and a suggestion.

Complaint- The 2 or 3 deep fried overblown tracks that exist in the rotation. Painful to hear when going from normal audio in my headphones, useless as training data, and useless as a captcha/overwatch. These tracks come back repeatedly, as well as some full on white noise. I just end up muting the tab, but it would be nice to be able to report the actual garbage tracks to get the removed for something that could do the job better.

Suggestion - look at my account language preference when feeding me tracks to "assess the lyrical accuracy" of. I literally cannot read japanese/korean, so I can basically only listen for words vs no words. It takes 15 seconds of listening before I can even press the skip, if I choose to do that vs just clicking "neither" and getting credits for my time.

That's about it, being subjected to 30 seconds of painfully garbage output, or being forced to listen to 15 seconds of something i can't understand before I can skip to a task I am capable of.

Edit - for some clarity, these overblown painful on the ears tracks have come up as frequently as 4 times in 5 total tasks. The exact same track for multiple different tasks

[spiral bronze]

Honestly the entire filter list could be simultaneously simplified and expanded by just making each type of track in the filter list have two columns: include and exclude. Then let people tick the boxes as they see fit.

[remote leaf]

Feature request: if I want to learn how to play a song that Suno generates, it should be able to create the sheet music or guitar tablature for the song. Last week I generated a song that I LOVE, and I want to learn how to play it on the guitar. The services I have tried so far for figuring out the chords and notes are failing me.

[knotty rain]

As others have pointed out, this isn't always suitable. Most of my own tracks feature a horn section, and I'd like to be able to place the trumpets, trombone, and saxes precisely, but I'm OK with the drums taking a stock stereo spread. So, having options is a good thing. 🙂

[knotty rain]

I generally precede it with [Intro - Spoken]. Then the next section is usually [Verse 1 - Sing].

I kind of take it to an extreme with this track, though:

https://suno.com/s/wNzei0TM0xsMrl4b

[pale imp]

I'll try that.. I don't usually put tags in my verse headlines anymore, but I'll look into it.

[scarlet wigeon]

Hi,
I realize that when collaborating, or simply when opening up a track
for a remix or a project with a partner,
the track in question must be made freely available to everyone.
Given this, wouldn't it be possible to grant access to one (or more)
individual artists without giving access to the entire community?

[last elbow]

What was the point of removing our ability to arrange our profiles? Would very much like that back

[blazing minnow]

100% with his opinion, it was a perfect way to presents our profiles, and now everyone as the same ...so why?

[cunning anchor]

Not entirely sure if anyone has noticed the beat or tempo for the early versions of SUNO 4.5 and up, are better for creating music in the metal genre

[mighty creek]

Man, The playlists on the web, are super laggy
I hope the Playlists Improvement comes
But also mobile we can have more than least songs we max have
On playlists

[haughty dome]

Now that you've moved several items into a "Manage" set of options, can you please add "Manage" to the bulk selected set of options, so that items in a Playlist can be moved into a particular Workspace in one go? Thank you.

[haughty dome]

I'm curious. What is the intention of inviting the Creator to Create a New Playlist for the purposes of Inspo? Either it's for some future plan, or it's redundant.

[clever harness]

In case you get to the screen and don't already have a playlist you want to use for the inspo

Or imagine you clicked that and had created no playlists at all, and only learn about both the feature and that playlists existed when you clicked on Inspo

[paper stirrup]

If any admin is online, I'd like to leave a suggestion here: in the next update, please add a function to use two Persona voices for duet parts.

[haughty dome]

mmm, a tenuous response. You can't create a playlist right there, though.

[clever harness]

It creates a playlist behind the modal. A beyond suboptimal user experience to be sure.

[hot violet]

Make it so [male vocal] sings in a male's voice and [female vocal] sings in a female voice and follow it.

[wintry perch]

Just want to let you know Suno... This is the point I am at.
🤣 🤣 I am sat on my hill. FIX THE SEARCH FUNCTION! It annoys me daily!

[slow walrus]

Why I'm not allowed as the channel owner posting a link of my own song to a comment I got? oh come on suno eh 😅

[west elk]

You are asking for support to scrape the site. That's generally considered a d!ck move among web developers. Better would be to advocate for an API to allow you to access the information you want. Scrapers are fragile and likely to break any time there is a change to the site.

[wintry perch]

No You misunderstood I asked an ai how hard it would be to make a search tool. I do not understand this stuff and then took a picture for the devs. To be like apparently this it is this hard. (I am not intelligible on these matters which is fine. I am in no way advocating for anything horrid.) It was half in jest and half in frustration if the tool it suggested is doing something awful then i shall rectify that. I took it down just to be safe.

[quick sequoia]

make "pinned songs" not transfer users directly off of the artist page! no idea when it happened, but pressing play on a pinned song sets a queue of 1 song only, leading the autoplay algorithm to instantly bounce off of the creator's page.

[hollow steppe]

You removed the ability to see how many times a song has been played on the web version in the list view. That’s really disappointing.

It was a very useful feature for identifying the songs I genuinely like most in my own library. For me, the fact that I keep coming back to a song again and again is one of the clearest indicators that it’s a great song.

I wish you can get it back, and better, to search or rank the versions of a given song based on this indicator

[west elk]

Yeah, I could tell. It was clear there was no malicious intent, just genuine "I don't know any better." This is the problem with AI tools. We can ask them to do anything, but unless you have the background to understand the answer, you won't recognize a bad solution when it gives you one.

[hard tide]

The play count is still there next to the play button for each song

[wintry perch]

That's why I normally don't use them at all. But I thought it would be incredibly funny to outline the steps for SUno so they could impliment the search bar correctly on their own site. (like it used to be back in 3.2 before they broke it.)

[barren wind]

this pls ty

[delicate cedar]

Please see #ticket-help for account related issues.

[west elk]

I gotcha; I've been there too. For what it's worth, I don't think it's development talent that's lacking, but possibly a lack of design direction. I worked with one of the devs awhile ago to improve the search feature (adding the ability to look for disliked songs, and a few others). They seemed quite capable. Based on that experience, the best thing you might do is provide "when I search this, I get this, but expect this" or possibly even just "I want a way to find this result."

[uneven blade]

I do have one suggestion I would like to make. This one I really need. Almost can't live without. But currently if I want to speed up a song it set increments using studio. A a slider bar is much needed. And an option to keep pitch or not, that could be a toggle.

[wintry perch]

Oooo Okay!
Search Bar& Resource Implementations to help with the ability to sort things for both Trash, Workspaces & Library!

1. in 3.2 Search used to search lyrics and titles {Search currently only searches titles I remember 4.0 breaking this feature and it has never returned} typing in Blood, should bring up all titles with Blood in the title & lyrics.
2. Add in ability to search keywords EDM, Pop, Rap etc.
3. Ability to search by Voice (because you can currently click on voice but it doesn't always catch them all)
4. Ability to search by stems, vocal only/ instrumental only
5. Ability to drag and drop and move your songs around. I would love to be able to move songs i want closer to the top, and group songs that I might have had an idea come in between and ruin my workflow. Having this for all library, trash and workspace.
6. Search currently for some reason searches only when a word is fully completed or one letter is typed. It does not active update to show as you are typing (this is slightly deterred especially if you start misspelling something unintentionally) like Poison being spelt Pioson, it won't pull up anything with that word in it. Recognizing misspellings would be helpful but not high on the list in needs comparatively.
7. thank you!

[pale imp]

Hey.. I can't seem to get it to verify my voice.. not sure what's wrong.. just picked up a new microphone.

[clever harness]

#1486542611475992749

[wintry dew]

It would be exxcellent if the video export exported the video cover when it has one..

[plush talon]

What did they break now as I am unable to create a voice out of a song as the option in the dropdown menu is gone so where is it moved to now?

[glass barn]

Keep collaboration updates in a separate channel 80% us only use it for looking at new features

@tacit idol a scam to delete here

[tacit idol]

Got 'em 🫡

[fast coral]

Why can't I upload my self-made beat? It says it's similar to existing copyrighted works.

[frigid bough]

This is the stupidest and most annoying update you've ever created. Thanks for making our lives a little more difficult.

[delicate crypt]

Probably a bug, I can't upload HYPERDRIVE by Riot Shift, even the isolated drop. I even tried a non copyrighted song like Into The Blue by J.A.K, gets blocked

[modern latch]

Suno should strongly consider stopping these unpolished updates. Any new algorithm or feature should be rolled out to a limited testing group first, and only pushed to the entire user base once stability is confirmed.

[delicate crypt]

Yeah

SUNO AI is kinda crap now

[upbeat mantle]

Just a friendly reminder have you read trough the ToS?

[delicate crypt]

No, is it good?

[upbeat mantle]

Well, you know that putting copyrighted songs into Suno breaks one of its rules and violates the Terms of Service.

[gloomy schooner]

@delicate cedar so we have a box that's only displayed once (well, I can't see it anymore) that states that covers of public domain material isn't allowed, but the TOS that state that they would be allowed, so which one is right?
(and yeah I've emailed Suno - but I'm not really expecting a reply)

I think it's something the team here should at least know or agree on

(and I'm not talking about the false or real flagging, nor the ones doing covers of copyrighted songs)

[gloomy schooner]

Even this server is false-flagging lol.
What could possibly have me TO for 10 minutes in this message? I see no profanity(??)
(and yeah it was a bot, the message was immediately deleted)

crippling

oh

it's l*wyer lol

you have l*wyer in your swear list? 😂

I guess it doesn't apply to everyone:

[hybrid crane]

The TOS and the popup are not actually saying different things. The TOS is broad and says you must have the rights and permissions needed to upload and let Suno use the material. That does not automatically mean “public domain covers are always allowed.”

The popup is more specific to how Suno currently chooses to operate the platform. They can still block certain uploads, models, or use cases even if the TOS language seems flexible enough to allow them.

So the real answer is: Suno decides what they allow in practice, regardless of how people interpret the TOS. The TOS is written broadly on purpose so they keep that discretion.

That is why support replies often sound vague or inconsistent. They are avoiding making hard legal promises while moderation and operational rules can still change at any time.

[gloomy schooner]

still nope, it states "and/or"
also, alex just wrote this, which is halfway (your own cover of a public domain score would be fine, but not a public domain recording of a public domain score? Yes you can find, on YT, pre-1926 recordings. I've used one myself)

[hybrid crane]

You’re reading the TOS like code syntax, but platforms do not operate like that in practice.

“And/or” does not create a guaranteed permission structure. The important part is still “rights, licenses, consents, permissions, power and/or authority necessary,” and Suno decides what they consider sufficient for use on their service.

You keep trying to find a loophole or contradiction to argue against, but there really isn’t one. You just disagree with how much discretion they keep for themselves.

Anyhow, I’m asking you, friendly but firmly, to move on now. You’ve made your point and you’ve been answered.

If you disagree with the answer, feel free to contact [e-mail] directly. Continuing to argue it here will just lead to further timeouts, which I’d rather avoid, but it is part of my job to enforce.

[gloomy schooner]

I agree on that, Suno can change its terms at any time, however you can't keep telling users to go read the TOS and then say that it doesn't matter because regardless of the TOS Suno can still do whatever it wants.

Besides, the customer is generally well protected in our countries (and very well in Europe), you can't just start crippling features/access that users have paid for in advance, for at least the duration of the contract, that contract goes both ways, Suno has agreed to provide a service for a duration, and the user has agreed to follow the rules at the time he signed.
Well, many have lost in court for that.
Also, I assume you're not gonna tell why you TO'ed me again?

And yes I have already contacted suno's support

[hybrid crane]

If you disagree with the answer, feel free to contact [e-mail] directly. Continuing to argue it here will just lead to further timeouts, which I’d rather avoid, but it is part of my job to enforce.

Moving on.

[gloomy schooner]

yeah just did, along with screens of this conv

[ornate bison]

I have a question. I generated a song while using trying Suno free version, couple of days back. I subscribed to pro plan today . Can i use that song from the free version for my instagram business page (commercial use).

[west elk]

No. However, if you make a new song, such as a cover of it, with the pro plan, you can use that version.

[ornate bison]

yupp i tried that. couldnt get the same result 🙂

[west elk]

Yeah it's never the exact same

[ornate bison]

it would be great if new users are given that option. otherwise how can people try out the software

[west elk]

I agree it would be great. But it is a false dichotomy to say "otherwise how can people try it out." You can try it out just fine without being able to use it commercially.

Try a cover with audio influence at like 60 or 70 %. Copy the style tags from the original song into the covered style section. That usually gets me something really close.

[hybrid crane]

You could have. Suno always gives 1 song you made on your free account, but then you had to do it by trying to download it. At that point you would have received a popup stating that if you upgrade, you can get the rights of that one song.

You still can get it, but you will have to contact [e-mail] to get that arranged.

[hybrid crane]

Covering a song does not give you copyright ownership, by the way. If the original track was generated under Suno’s free tier, Suno still holds the rights tied to that output including the melody. A cover is still based on and recreates that underlying composition, so it does not suddenly become your copyright just because you re-recorded or remade it. The underlying rights still trace back to Suno.

[novel compass]

ummm.... I'm sorry, what? This better be a glitch.

[young fractal]

WTF?!

"Your lyrics contain copyrighted material. Please change it and try again."

Weder "Stille Nacht", "Heideröslein" noch andere Uraltlieder haben ein Copyright...

[tepid trench]

This is being looked into. The projects still exist, there's just a problem displaying them at the moment.

[novel compass]

😮‍💨 ok thanks.

[spiral bronze]

This may seem like a weird or useless suggestion, but hear me out.

So we can obviously type something simple into the prompt window, like "classic rock" for example, and Suno will figure out how to turn that into a full song anyway. Or, before we hit generate, we can get Suno to enhance the prompt and then it'll do that song as described. But, if you generate a song with only "classic rock" as the prompt, there's no way to get Suno to look at the song is actually made and then tell you what's in it. I think it'd be cool if we could hit a button and Suno will generate a prompt from audio its already made. This doesn't have to be anything different from the tagger it uses on uploaded audio, but it'd save a step if you just want to see how it would describe the track.

[novel compass]

I get a similarly silly error about using a producer name when my lyrics have “you’re dead” or the word “phosphate” in them.

[spiral bronze]

It's good to hear they care about all those phosphate producers out there. That's a big export industry these days.

[novel compass]

lol. I mean, without them “you’re dead” it seems.

[tepid trench]

Should be working again - may need a page refresh

[novel compass]

I’ll check it when I get a chance. Thank you.

[golden acorn]

Hey folks!. I feel the need of a feature where we could use the auto-generated title as a {param} within a string, to help easily classifying songs, so here's what I had in mind with this draft mock-up. Thanks!

[hybrid crane]

Please do not ping people from the Suno team. Posting your suggestion is enough. Staff goes through them daily. Thank you for your cooperation!

[golden acorn]

Sorry about that, edited it. Thanks

[cobalt mortar]

Any chance of easing the thresholds of copyright matching? Low enough threshold and anything that vibrates air matches. XD

[west elk]

So your proposed feature is... to include a tag in the title? You can literally do that now already. Just include a tag in the title when you write the title. Granted, it's not automated, but most songs have so many tags, it would be a mess to try and pick anyway, plus, you don't exactly have to make song titles very frequently compared to how often you do anything else.

The Suno team doesn't seem to have a great abundance of development time. I'd rather they spent that precious time on something that we can't already do easily by hand anyway.

[atomic grail]

In my work sometimes I need to keep drums out entirely. Could you provide a way to do that?

[vital jay]

This is also an overall suggestion but an option to create a duet would be stunning.

[golden acorn]

Well, granted, this is more like a "nice to have" feature. It does help solve having a form of "unique memorable" name while also allowing (as you called it) tagging, so for instance, you can convey the mood of certain audio songs directly, and it seemed a rather low hanging fruit. But I do have my own scripts that are able to do that post-generation, so, priority-wise, I definitely agree with you. Let the heavy backlog make its way.

Instead, maybe a better option, now that you've said this, would be either:

1. Expose some form of API to allow users to write their own local plugin implementations they can use for themselves;

2. Expose a part of the UI source code so users can contribute themselves and make PR requests.

I wouldn't mind sharing some of my own time for features like this one which almost never make it out of the backlog since others always have higher prio.

[obsidian peak]

Suno's infrastructure is expanding; the company has literally hundreds of millions of dollars. I demand that the character count in the style box be increased, please. 1,000 characters are not enough.
Three new models have already passed, but the limit is still the same

[tepid laurel]

Its super annoying that my songs keep cutting off the last verse at the same 5:30 Mark,

Yet it didn't sing faster to cover everything..

The annoying thing is that I KNOW you have a Duration setting, all it needed was 20 seconds more.

EXPOSE THE DURATION SETTING.

And what about maxMode? Lol

[violet sierra]

So I see I can add a caption to my highlight pinned song, but it already has a caption on it. Why not automatically use that? Why make me add another one? How intuitive.

[lethal pivot]

Does anyone know if A.I. Music Generation (As in all of them, Including Suno) is Injecting Nefarious Frequencies into the music?
I tried two songs in an unfinished website and it came up with 20 ISSUES.
Some frequencies can cause Nausea, Stress, Anxiety, Depression etc... Some modulated frequencies can send messages to your brain.. But you hear nothing.

[violet sierra]

Big ups, thumbs up on the My Taste section. It has auto-generated the description of exactly the stuff I like. Seems like it can apply that to what I make, but can it also be used to find stuff in Explore? Like in a 1-2 click way? This could be a game changer!

[kindred hull]

I also have experienced that issue, but primarily using personas/covers from songs from older releases of Suno. Could that be your issue?

The feature I miss most is very simple: the search bar from the web does not allow using more than 1 word. For instance, "god" gives me as result "god in the groove" (in my library, of course), but if I search for "god in the groove", I get no results!

[tepid laurel]

nah - its a new persona. it just has some pre-defined maximum duration of 5:28 or whatever even though that was never agreed or negotiated

and extends of course are still broken since they include Previous lyrics in the extensions prompt - so if the first part was 5000 tokens it breaks its brain... because for some reason the model needs to see all the previous lyrics in order to infer what the song should sound like next >_>
(it doesnt actually require that but yet here we are)

[obsidian peak]

Why my suggestion didn't get any like guys? You don't want better quality??

[clever harness]

Haha... there are only a few experts on planet Earth doing AI music at the moment. The next time I see him I'll let him know he's wrong.

[tepid laurel]

I tested it lol. Remove Part 1 Lyrics -- the extension worked great. Audio is enough context - the model already can hear and understand the meaning of the spoken vocals in the audio without needing to see the semantic verbatim text of the old part.. the audio And the New lyrics were more than enough to get it right..

but when part 1 is 5000 characters of lyrics -- and the max prompt length is already at that point.. so it can't "extend" anything unless irrelevant contexts are temporarily discarded.

But to have to delete Old lyrics from Part 1 - then add them back when the extension is done -- so when I make the Full song it has both lyrics.. that is some thing Suno should be doing automatically as a POST process -- Not During Generation, feeding both lyrics into the generation model ..

[hybrid crane]

I am not sure what you are trying to say but that is how it for me always worked. You always remove the old lyrics. People who extend with full lyrics without removing parts that are already sung never worked. And is logical imho. For me personally extend works like it did with v2. 🤷‍♀️

In other words.. I don't have that problem to begin with

[clever harness]

What I was able to reproduce was as follows:

Setup

1. Use all 5000 characters "Hello I'm Alex Ayers" Song A
2. Do an extension wherever: "Hello I'm <B>" Extension B

Expected Result in Extension B:

Extension only has "Hello I'm <B>"

Actual Result in Extension B:

Extension has a few lines of "Hello I'm Alex Ayers"

If before I do the extension B  I wipe all displayed lyrics from Song A. I get my expected result.

[tepid laurel]

^^^ exactly

sorry my image example has sleep deprivation echoes.

[tepid laurel]

And yeah I completely agree - and thats what i thought i was doing at first.. only giving it the next lyrics.. But it turns out it still sends the old lyrics too even if you dont put them in the prompt - you have to go to the Song's Info and actually erase them from the Link -- store them somewhere safe temporarily, and then later put them back into the song page..

[hybrid crane]

Weird.. I extend frequently and never had this problem.. But my lyrics are far below 5000 characters...

[tepid laurel]

Yeah i think it only becomes noticieable at that point because the model is smart enough to just ignore what it knows is already in the audio i guess. but at that 5000 limit it can't really even see the new lyrics yet -- since they're way at the end.. so its like "I guess you want me to sing this all again? OK"

its outside of its horizon of sight basically so it can only assume you wanted to loop it again ..

Possibly if the order was reverse so the extension lyrics come first, and then Part 1's are looped back at the end -- that would actually solve the problem and still have the benefits of the loopback stuff.

[west elk]

1. I would like more space in the style area. Agreed.
2. Your claim that "Three new models have already passed, but the limit is still the same" is specious. We used to only have 200 characters for style.

It's one thing to make a valid request. But it undermines your legitimacy when you make "supporting" statements that are false.

[latent compass]

Hello, it seems not possible to make 2 remixes with the same song with the upload audio feature, how to correct it?

[west elk]

Is this a request or a suggestion? Or are you asking how to do something?

[latent compass]

well, i'm not sure, but when i upload an audio already remixed, it's refused by suno, is this a bug?

To be precise: even if I delete the previous remix or the uploaded audio from my library, Suno still rejects the exact same file when I try to upload it again for a new test. Is there a cache/signature block, or is it a known bug?

[hybrid crane]

It's not a suggestion, that's for sure.. Try the #🪲┃bugs-web

[umbral oxide]

I’d like to suggest a few improvements that I genuinely think would make Suno even more powerful for creators who build detailed, cinematic, or long-form music projects.

1. Increase the lyric character limit from 5k to 10k. A lot of songs, especially cinematic tracks, story-driven songs, progressive songs, rock operas, EDM hybrids, and emotional narrative tracks can easily go beyond the current limit. With longer song generation now possible, it feels restrictive not being able to fully use the available runtime because the lyrics run out too early.

2. Increase the style/prompt limit from 1k to 5k. Many creators rely on detailed prompts to shape vocal tone, structure, pacing, transitions, atmosphere, instrumentation, production style, and emotional progression. More prompt space would allow users to guide the AI more precisely, especially for advanced productions and genre blending.

3. Add support for structure commands like:
   [pause]
   [split]
   [stop]

This could dramatically improve musical control and allow more cinematic arrangement behaviour. For example:

• pauses before a massive drop
• sudden silence for emotional impact
• beat cuts
• dramatic transitions
• vocal interruptions
• cleaner section separation

These controls would make generated songs feel far more intentional and dynamic.

4. Add a song length slider. It would be useful to choose a target duration before generation, such as:

• 30 seconds
• 1 minute
• 2 minutes
• 4 minutes
• 8 minutes
• longer in the future

This would help people creating:

• YouTube intros
• trailers
• TikTok clips
• short themes
• full-length cinematic songs
• extended workout mixes

Please don’t let this feedback disappear into the void. I’d genuinely appreciate some indication that suggestions like these are actually read and considered. A lot of creators are passionate about using Suno creatively, and features like these could push the platform even further.

Thank you for all the work already put into the platform.

[clever harness]

What's your account? I'll try a patch and flag you into it.

But likely not shipping for like two weeks if I'm being honest

[wispy matrix]

1. I replaced different sections in a song but theres no way to control blending in between sections...even if i select the section in between and try to redo it, it just doesnt work.
2. I want to control vocal effects better not just save voices that suno created. Yes I have uploaded my own and even described vocals as best as I have learned but there should be a way to apply those effects to my saved "voice" for each song prior to generating. Otherwise I get stuck with vocals over the music that just sounds like its too upfront or something.
3. I want a way to remove things from a song for example I want to highlight a lyric where the effect takes place say echo and remove the echo.
4. Expiremental is fun but how do I know how to do it again? I want suno to Describe the style so that I can use it on a future song.
5. I would like 1 button to like equalize or average out the songs volume throughout the song so that no part has a sudden loud shift.
6. A way to remove any parts of the song that are silent for more than x amount of time like if for some reason there is an awkward pause of 2 seconds in the beginning or end and wherever else in the song where is the button to remove all2 + seconds of silent?
7. A genre checker...because sometimes i have remixed so many that I kind of dont know what genre my song should be considered. Is there an expiremental genre lol.
8. It would be cool if somewhere there were options of style descriptions that could be used along with a sound of say a 10 second melody that is not the definition of the style and then a 10 sec one that is so we can learn by hearing because some of us learn by hearing. Like what is "four on the floor" ? My adhd brain does not like that one at all

Oh and one more...lets say I only like the middle part of a song I want a dual ended extension option yes I want to pick the amount of time each way and ill write the lyrics...suno can take care of the music so that it agrees with the middle.

10. I imagined being able to save any cropped intros or endings somewhere and then having an option to apply that to any song as an edit . In doing so suno could blend the bridge well?

Finally, if there is a way to write lyrics in a way that tell suno to hold the vowel in a word let me know because I want intros like that

[brittle wedge]

I WILL PAY $$$ FOR DISTRIBUTION TO GLOBAL STREAMING SERVICES

here is my suggestion, I would like to see SUNO as a leading distributor of AI generative and assisted music production. I am already a member of AWAL, Bandcamp, Ditto and SoundCloud and I already pay in one way or another for distribution. I’ve also been a writer member of PRS since 1988 and I’m a member of PPL.

I’m a musician and I want to create things. I rarely think about the mechanics of copyright or publishing and I want this handled for me. It’s clear that currently streaming services are thinking about this a little too much and the whole distribution thing has become over complicated in a saturated market.

I want a Premiere Plus account that works like a music distribution account WITHOUT the complexity of a distribution account,

I want to do my own editing, mixing and mastering in ABLETON, or LOGIC.

My Meta Data should simply read writer Mark Tinley assisted by Ai.

I don’t want all the extra steps involved in getting my track to market.

I don’t need to hide the fact I’m writing using Ai to assist me, it’s become a viable music production tool.

I don’t want to spend ten times the amount of time it takes me to make music on mundane admin.

Make SUNO a label.

[brittle wedge]

I agree with you about the 10,000 character limit, I’ve run into the problem a few times myself.

[hollow frost]

Maybe Increase the number of custom models that can be made at once

[knotty rain]

I've been wanting that for a while, using the Voices option. Or a trio. Or an ensemble song (a large group handing off vocals, like "We Don't Talk About Bruno").

[vivid cipher]

i have paid the amount for suno as a pro plan on 29th april and now they have been deactivate my account what is the reason suno tells to active the subscription

please give a support as soon as posible

[quartz hare]

Better embed player for music and Suno links on discord. IE the ability to play the song/hook/media without redirecting/tabbing out

Maybe even a Suno capable music player bot

[median phoenix]

Please make it possible that the AI can generate songs in any language not depending on the selected language

[hard tide]

It would be nice to have the Filters expanded so you could have the Library group alphabetically or by voice (currently the only option is voices with newest/oldest sort).

[golden moth]

Discords chill chat atleast needs to sllow non suno links to our music

I s hould not be req to share from there or send file in chill this is abusive moderation

[delicate cedar]

This is Suno.

[golden moth]

And your point

Many ppl myself included move there songs off suno to polish it and add it to better detailed sites

So in chill chat it should let us post thise links snd not forcing us to be unable to share our full work

[kind grove]

For everyone who needs more voices than just their own, or prefers not to use their own voice, but still wants a fitting and unique voice for their song.

Now that we have the ability to use our own voice in generations, the next logical step would be a voice generator that allows users to design their own generic/custom voices. This would be a huge milestone, especially for story-driven music projects that rely on recurring characters with consistent voices.

Is something like this currently being considered or planned by SUNO?

The voice generator could either work with sliders for different vocal parameters such as tone, pitch, warmth, breathiness, raspiness, age, accent, vocal strength, and emotional character — or alternatively through AI prompting. The result of the generator should become a usable voice within Suno’s voice system.

To make these voices even more useful, it would be great if multiple voices could be selected simultaneously during generation, allowing true multi-character or multi-vocal songs. It would also be very helpful if these voices were available inside Studio during vocal generation and editing.

[kind grove]

Hi,

A few weeks ago, I tried the Own Voice feature. I recorded myself singing one of my songs and uploaded it to SUNO as a voice. Technically, it worked quite well, but I noticed an issue:

I am absolutely not a singer. Apparently, I ended up singing with my speaking voice rather than an actual singing voice (at least according to feedback from a musician community). I think it could still work as a character voice — maybe for a clumsy or naive character — but not really for serious vocal generations.
https://suno.com/s/pYMsSwszsJRo7Bde

It would be very helpful if voice training could also work by simply reading a random text provided by the AI, allowing the system to estimate or generate a possible singing voice based on the user’s natural voice. That way, even non-singers could achieve cleaner and more polished results.

This could also work as a potential workaround for my suggestion in #💡┃suggestions-web , especially if live voice modulation during recording were possible (assuming that would be allowed).

[hollow locust]

Give me the option to NOT HAVE a 30 sec. intro. I hate it. 10 ~ 15 is all it needs to be. 30 is overkill.

[clever harness]

Why not use an extension at the point you want to change up the song?

[clever harness]

You can type any lyrics into the lyrics box, how are you generating

[low harness]

What i would like is that the extend is corrected. Because is the best way to get more than 1 voice and sometimes, i hate that the song extention just goes to 4:00 every Extend. ( And you don't want to get a 4:00 while you just want it to end)

[limpid gull]

Could we get a vocal mixer? I hate mixing my own vocals but Id like for the song vocals to be my own without having to try to recreate the song again using the personas and having suno make it all different. The personas are cool and all, but they aren't the same as replacing the vocals of the song with my own it more makes a song like it but not the one that I perfected.

[crisp ember]

The mix of languages all over SUNO is becoming worse. Can we have language selector?

Just some "highlights"...
pic 1: "Stängel extrahieren" - this does not make the slightest sense! Neither in "real" German nor for musicians (extract STEMs). And this is just the worst from the Account page - the mix of teo languages is really annoying. I would prefer English only...
pic 2: Explore page: mix of languages
pic 3: all in English besides "Stimmen"
pic 4: recommended styles... I don#t know how to say, but some of the underlined ones do not make sense at all, some don't ring a bell, some really are German words but do not have any meaning in musics, some look like an error

[hollow locust]

Because I want the lyrics to start at 100ish sec. Not 30-ish sec. v5.5 seems to he hard-coded to have a 30-ish sec long intro, no matter what I do. An extension, in my experience, won't fix that. And I can't really extend from 00:00. So... that's why.

[knotty rain]

Totally agreed! I often end up going into the Editor and cutting out measures, if I can even find places where stuff can be cut.

[spiral bronze]

Would be pretty cool if we could set up "cues" for our tracks before genning them. Like I don't want to load up Studio every single time I want to edit, but like people are saying about intros being too long... well, I very frequently have the opposite issue. I'm trying for atmospheric, I want an unreasonably long intro, but I can't get it because extend only works on the end of a track. So if we could set up our segments with timecodes and toy with them, I wouldn't mind spending 2x or even 4x the credits. Means less luck-of-the-draw and less spam of "Close, but no cigar" gens cluttering up my workspaces.

[hollow locust]

Exactly! Especially with that "if I can even find places where stuff can be cut." part!

[short narwhal]

it would be very useful if new tracks that have the same title would be numbered. It is really hard to find the track later on when all tracks have the exact same title

[wise urchin]

It would be great to have seperate prompt boxes for vocal description and instrumentation

[modern latch]

Suno, you sell services, we are the customers. If the service doesn't live up to the description, and customers complain and get angry, what's wrong with that? And furthermore, while there are tons of errors, Suno is even more aggressively running advertisements.

[clever harness]

Perfect. Thank you.

[modern latch]

Ultimately, what Suno is doing is showing disrespect to its customers.

[quasi goblet]

Before downfalls comes in arrogance, rude behaviour and lack of empathy towards loyal customers, Suno is no different!

Once a competitor emerges lets see on how these bots behave with us!

[cinder lava]

Personas/voices aren’t able to be used with uploads. Although I understand the concerns- this is a huge opportunity lost.

The best workflow I have is downloading my suno song, uploading it back to suno and having the llm re-evaluate it then remix it.

The added step introduces complexities and hallucinations that give much more unique results and especially voices when playing with the weirdness slider.

Suggestion is a metadata tag added to downloaded songs that can be confirmed by the llm to lift the voice restriction.

The new copyright checks are ridiculous every melodic track, even songs generated by suno are flagged as existing works of art, destroying my previous workflow.

[violet sierra]

Bingo! Also maybe being able to appoint a voice to a Name text or use a specific voice/persona. Duets are still not very stable at all.

[tired smelt]

I don't know if this has been suggested before, but here goes. For playlists, it would be nice to switch to the songs that has been recently added. It's by default (for me) the older songs are on top, so a button to switch to the newly added ones would be great to have.

[ionic glade]

曲のアップロードの規制を消しなさい！
コレは人を侮辱してます！
皆怒ってますしsuno aiを皆辞めますよ。
こんな酷い状態を4日も放置してるのは運営としても、どうかと思います。
即座に規制を消しなさい運営として信頼を失ってますよ

[mighty creek]

Playlist Update! for web, which stops lag

[remote trench]

An upload feature that actually works without false copyright flags would be nice 🙂

[tepid laurel]

Just saw this! wow im late.
https://suno.com/@military_technical

[west elk]

Are you aware that you can add musical direction in the lyrics section? Enclose it in brackets, just like a section tag.

[west elk]

What are you actually complaining about though? This is just "We're angry, and you should listen!" There's no actual substance to what you said. You want to see change, then use your words and explain what your problem actually is or what change you want to actually see. But "Yeah, we're mad and we're not gonna take it anymore" statements mean absolutely nothing and accomplish nothing.

[sweet zephyr]

Yep, I uploaded a tune I wrote 20 years ago in Acid Pro and it got flagged as existing works (copyright) on SUNO which is BS.

[delicate cedar]

See #🚀┃updates it's being worked on. Fill out the form and submit it please.

[ripe summit]

I completely agree with copyright checks, but since the technology is already here and it's so convenient, I wish there were a way to search through my own generated song library by playing the melody on a piano or simply humming the tune.

Many times I remember the melody, but I forget the generated song name.

[lusty blaze]

Does Suno have a way to make songs a certain length? If not, it should. I need a filler piece of music 55 seconds long. Suno cannot seem to handle that. That would be my suggestion for an upcoming update. Specific length for songs.

[wide cradle]

There should be fix lyrics mode. Exact audio and vocal just replace lyric only. Cover no matter how I try with various config, it always drift somewhere. Thank You

[rough imp]

谁可以帮帮我，我是中国用户，现在我用Google账户登录就是提示这个，我不知道问题出在哪里，也联系不到工作人员，我的英文很差，所以打了中文，希望得到帮助，谢谢🙏

[modest mirage]

/support

[rare obsidian]

SUNO SUPPORT HELP

For support assistance, please contact the correct team by email:

Billing and subscription issues: [e-mail]
General support and all other concerns: [e-mail]

If you need help understanding who to contact and what information to include so the process goes more smoothly, feel free to check the guidance provided in the #support-guide channel or start a private thread in #ticket-help for assistance.

Please note that Discord is not a secondary support channel and will not speed up existing support requests. Support is handled in a queue, and requests are worked in the order they are received.

[modest mirage]

The only way to contact support.

[rough imp]

我给[e-mail]发了邮件，中文描述的，不知道能不能理解我的意思，我登录了iOS移动端还开通了付费订阅

[velvet horizon]

So, yeah. Just got this error message. Song was acapella beatboxing and melodic scatting. There were no lyrics, not a single word was uttered, just mouth sounds. It was sadly a direct recording single take and I lost it, not happy right now. I get trying to prevent people from uploading copyrighted material but this is going too far. I hope I don't get this error again, considering how much money I spend on y'all per month. 😠

[modern latch]

I've told Suno before, but they just wouldn't listen:

1. Before updating anything, you should announce it. If you don't announce it, then test it in a localized area; if it's stable, then update the entire server.
2. With updates that cause so many problems like this, you should uninstall the update.
3. Since this new patch came out, there have been tons of problems: interface errors, terrible music output, copyright issues, etc.
   They are a large company now, but they behave like thugs on the street.

[silent wren]

Would love to suggest something to create group of workspace, like if we could group workspace into folder/group

[balmy comet]

I can't find another category, that matches my request:
It's a shame, that all uploaded content is normalized to -16 LUFS.
If you download a track to correct some aspects of it, it would make sense, you can at least reupload with the same loudness parameters, than the original generated track.
Example: https://suno.com/s/JeuqaPCj5NYctqJp

Look at images, first is WAV download after generation, second one is reupload. Doesn't make any logical sense to me, because: If i can cover the internal track with the already good loudness, there is no reason, why i would want to limit uploaded loudness. At least lower volume as already existing track in system.

[upper hollow]

Lately, I really want a refund because of Suno's copyright management. I deeply regret paying for the one-year subscription.

[drowsy cosmos]

Please add a language filter or a dedicated way to search for songs by specific languages to make global music discovery easier. Thanks.

[rough moth]

Please add an option to "Liked" playlist to exclude "my own songs". When I create songs I usually "like" the ones I might work on later, so the "Liked" playlist gets long. On the other hand, I'd like to to use the "Liked" playlist to listen to just some specific great songs created by others. This whish applies also to android and ios platforms.

[indigo rivet]

Same thing just happened to me. 100% original audo. This needs to be fixed ASAP!

[ripe summit]

I find it very interesting that most generative AI systems, such as OpenAI and Gemini, try to refuse generating copyrighted images or text. However, Suno chooses to filter the input rather than the output.

[winter island]

Hello. Previously, the application correctly sang several vowels added to the word: aaa.., and now it sings: a-a-a😂 Have you changed anything? Maybe they added a sign of lengthening the sound, but I don't know?

[hard tide]

Rolling back an LLM isn't the same as rolling back code for a website.

[opal thorn]

Hello, I'm not sure if this is the right channel, but I wanted to say that I would really like to purchase the Premier plan, but the price for users in Brazil is very high due to the dollar exchange rate. Does SUNO not intend to create pricing policies based on the users' region?

[clever harness]

[e-mail]

[worldly cargo]

It'd be cool if the items relevant to pro/premier licenses were properly labelled across the board, though I'm not sure if its just an out of place thing. Get Stems/Midi for example advertises itself as a Pro feature in the UI, when getting a midi download is actually restricted to prem accounts it seems

[clever harness]

I'll fix that

[ruby aurora]

There's a 2 hour timer in the other channel so I can't send the video there. Also your dms are closed too.

But here's a video of the lag

[clever harness]

Brutal, thanks

[indigo rivet]

Was there an annoucement about "uploaded audio containing copyrighted material" ?

[hybrid crane]

#🚀┃updates message

[hard tide]

You can choose V5 in the drop down if you want to use the old LLM. Personally v5.5 works just fine.

[clever harness]

Keep this playlist around, we'll be using it as an example playlist to test the fix

[knotty rain]

https://tenor.com/view/laughing-clown-buggy-one-piece-creepy-clown-laugh-evil-clown-laugh-gif-16525184346419446540

(Sorry. Couldn't resist.)

[hoary swallow]

I've been having issues today with not being able to upload due to "existing work". Funnily enough those are all tracked and absolutely no samples or anything. What's going on?

[hoary swallow]

ahh yeah must be this

but especially bad today, that i've been doing 100% recorded stuff, awful to go through, real roadblock

[clever harness]

Throw an example of a blocked upload into #🪲┃bugs-web

[hoary swallow]

anything i record

[clever harness]

One example will work if you want to share it

[hoary swallow]

not really kosher to share in public

[clever harness]

k

[worldly cargo]

Thank you for checking into it at all while bobbing and weaving out of the way of all the flak you are catching right now over this upload thing 🙏 Must be common but seems like an insane amount externally. Hero we dont deserve

[clever harness]

Appreciated. If I was concerned about angry users I'd not be here. There's value in keeping a pulse on the community (the good and the bad). These days I'm getting quite a kick out of being the bad guy to some people on reddit.

[mighty creek]

The Playlists are laggy, let say for example I check in web the playlists to use the inspiration thing, and check some songs... but it turns out let say I have 90 songs on a playlist in a web, well.... lag comes in. That it becomes harder to scroll down

Or even edit or something

A suggestion would be, to improve the Playlists... I guess to be less laggy, and more usable to use, since I guess I have lots of playlists

Cannot scroll my favourite playlists with my favourite created songs, without lags to my collections....for others to check, and even my private playlists...

[wooden lance]

i'd really love to see suno add the "collab" option with other suno artists just like on spotify or other streaming platforms, so a song made by 2 or more artists shows up on everyone involved profile

[wild nest]

Can you change how Reuse Prompt works? Often I'm reusing a song's prompt because I didn't like the first renditions, but Suno will make new songs with very similar intros, melodies, etc. It would be nice to tell the AI to create variance while still utilizing the same style.

[final echo]

Can we please have a slider that tells Suno the maximum length we want our song to be I'm sick of it generating 7 minute songs that cut off

[west mirage]

Can you add the clip share to tiktok via web please. There's no option for the browser

[brave siren]

is anyone else getting, like, just REALLY artifact-y results today?

[quick sequoia]

gift cards/codes?

seems like a win/win.

• I can attempt to share the premium suno models with people i think will enjoy the platform as much as i did
• gift cards = money for suno with a lower chance of using the actual computation for generating songs for every lost/unused gift card, and also a more specialized way for the users to attempt to bring in new people, far more effective than referral emails for 100 credits or whatever

[pine verge]

Hi @daring tusk .
Love how Suno is working atm. But! I have one request for future updates.

Right now with Voice Clones I feel I lack the ability to prompt in my lyrics when or what Voice clone should be used!
.I have 3 Clones atm from my own voice singing in different tonalities and styles.

So I feel we need a way to link to our Clones in the lyrics.

Example!
[Intro - Voice clone 1]
[Chorus - Voice clone 2]

Or.

[Chorus - Calm and soothing] (Href:"clone 1" sings in a storytelling fashion) There's one’s was an app called Suno. And now they made you all know!
(Href:"clone 2" dark and commanding tone) Just how great it feels, When you see your creation show up on someone else's reels!

Please make this a priority for future updates 🙏

[ruby aurora]

It's actually not my playlist. (But all playlists lag for me, and bigger playlists like this lag even more, including my own).

But some more info, the playlist doesn't lag at all for the first 2 seconds that the page loads, I can scroll and there's no lag at all. And then the second the text loads at the top, the description of the playlist, that's when it starts lagging. 1st image is there's no lag, then 1-2 seconds later in the 2nd image there's lag. Even before the side panel loads there's lag, it happens when the playlist text etc loads.

I bet this info is all you need to diagnose 🤞

[kind grove]

I really like this idea, especially for creators who work with multiple voice identities and storytelling concepts.

Personally, though, I think voice clones should mainly contain the voice characteristics themselves — for example:

• vocal tone
• timbre
• texture
• age impression
• vocal color
• accent/pronunciation
• vocal range

Things like singing style, emotion, delivery, or intensity should ideally come from the selected music genre and/or the prompt itself.

For example:

• cinematic emotional vocal
• aggressive industrial rap delivery
• soft intimate acoustic storytelling
• dark orchestral spoken intro

That way, one clone could stay flexible across many genres and moods instead of being locked into a fixed singing behavior.

But I absolutely agree that we need better control over which voice clone is used in specific song sections like intros, verses, harmonies, dialogues, or choruses. That would open up huge creative possibilities.

[pliant charm]

I think to be able to breakdown stem even more, like for drum, so you can get kick drum, hi-hats, and snare all separate and even guitar or vocals so that if you have 2 guitars or clean and unclean vocals you could separate them into their own stems

[indigo rivet]

Right now, the fact that a songwriter can upload his own work to Suno and create a radio ready audio track, is the most ethical and defensible use of SUNO. For people like me who use the tool like this, SUNO is a "session band" in a box. So it's misguided for Suno to be heavy handed in their approach towards creatives who upload their own original audio.

If a SUNO content flags a users uploaded original audio, it should have a human based appeal system. Anything less is going to erode trust. This is the problem with "platform" thinking. Case in point: X and YouTube have zero humans and instead are using AI to cancel and demonetize accounts. Once that happens, when a Suno killer shows up and it will, users will migrate to a new system. Treat your uses with respect now because it builds loyalty. Suno needs to be facilitating Apple computer levels of loyalty with their customers. Not in the future, but now when it counts.

[viscid minnow]

not sure if it's a bug or a feature, but ive noticed Suno's copyright has gotten very strict as of late. songs i have uploaded before to use for covers and mashups are now suddenly unable to be uploaded and it's flagged for copyright. can we please loosen the reins a bit or have the copyright only kick in when youre actually "publishing"/pinning a song to your account, or something along those lines? i'm not publishing anything i make on Suno nor doing this professionally for monetization, i'm only using Suno for fun and it's starting to get frustrating now that i'm trying to make an interesting cover or mashup idea and i get hit w/ the "matches existing work" pop-up. theres so many professional artists that make remixes and covers of existing work, so why cant we Suno users??? so long as we either dont publish it, or make sure the song isnt 1 to 1 and give credit where it's due if we do, whats the big deal? i'm just doing this as a fun pastime and, unless i generate straight from the AI itself and tell it to give me a completely original prompt, i'm getting more and more restricted w/ what i can and cannot make!

ah, i see the update now that this is a feature. my point still stands, this is too restrictive

[indigo rivet]

I agree 100%. The irony of Suno flagging a user's original audio should not be lost on anyone. People who live in glass houses should not throw stones.

[viscid minnow]

exactly. ive tried everything: official (like mainstream, what-youd-hear-on-the-radio), video game music, anime music, fanmade remixes posted on YT, and i even tried uploading a friend's song he made previously on Suno before this copyright trial started and none of them worked! this copyright AI is literally blocking everything you upload. the fact even original works ppl are recording via Suno is also being blocked means it's way too strict. do away w/ this copyright entirely or have it that it only kicks in when you try to publish it, not when you upload songs. the heck is the point of this tool if you cant even utilize it properly???

not to mention trying to set up this copyright blocker is very ironic for an AI tool. AI's whole thing is to copy/"steal" existing content, so having the AI block itself from using copyright material is very ironic. i get what Suno devs are trying to do, and it’s admirable, but like these anti-AI peeps are gonna hate on Suno and similar AI tools no matter what so why bother?

[clever harness]

ive tried everything: official (like mainstream, what-youd-hear-on-the-radio), video game music, anime music, fanmade remixes posted on YT

This is why the block exists today.

[viscid minnow]

studio-made songs are to be expected, i was only testing. but it blocks everything, and, again, i personally (and i'm sure other users too) dont monetize what i make

this is far too restricting

even ppl's original works they made themselves are being blocked

just as proof:

dont plan and never planned to publish anything. i'm just using this tool for fun and now it's borderline useless unless i just make AI-generated work

[wooden moss]

SUNO is in COMA right now, thats the TRUE, legal issues by lawers an Judges

I can't make covers of MY OWN SONGS, does not work, give me the same sound no changes

no errors shown

[split egret]

Add Vocals/Add Instrumental/Add Stem features all need to be fixed because none of them do what they say they do
Currently all of them just act as another Cover function, they generate a whole different song instead of adding something to an existing one like they claim to. If you have a function that's supposed to "add" something, that means keep what's already there and add to it, not make a whole new song
If for some reason you can't technically implement this (which you definitely could but let's pretend), then either rename those features to something less misleading, or just remove them since there doesn't seem to be any use for them

[split egret]

Also, and I have been pointing this out for a year, exclusions STILL don't reliably work and style instructions are STILL frequently ignored
The developers unfortunately don't care about anything we write in here though, otherwise they'd put some slight effort into making their product work as it claims to, instead of just looking for new ways to break it

[undone chasm]

Hi, I've been working on https://ehdarkmuse.pages.dev/. I noticed that the system seems to flag "capo" as if it were referencing rap or hip hop terminology. Could this be adjusted? "Da capo" is a well-established musical term, especially in classical and baroque music. Right now it tends to push otherwise baroque prompts toward hip hop aesthetics when using my app, which relies on very specific musical terminology rather than artist names.

Also, Suno seems strongly biased toward modern electronic drum kits. I'd really love to have more access to vintage-style acoustic drums, especially 70s-style live kits instead of electronic or EDM-oriented drums.

And another question: if I'm using names of musical techniques or historically established terminology like Bach-style counterpoint, Chopin rubato, Debussy-style harmony, Bartók pizzicato, etc., could those get flagged by the system? I removed them for now just to be safe, but I'd appreciate some clarification because I'd really like to use proper musical terminology in the Alyx Engine for my app.

From a music theory perspective, these are standard educational and historical terms rather than artist impersonation. They're commonly used to describe compositional techniques, harmonic language, articulation, orchestration, and performance practices. Blocking them entirely could unintentionally reduce musical precision and make advanced prompting more difficult for musicians and composers.

I also understand that Suno may not be heavily trained on certain classical conventions, so maybe some of these terms don't influence generation that much yet — but clarification would still be really helpful.

[tepid laurel]

**Make Songs that were made with Personas still show the Persona/voice option in the dropdown menu. **

Just because it was created with a persona does not mean the persona that was used is the end all be all.

Each song is a new thing, that deserves its own persona/voice (if someone wants). Currently it acts like if you used a voice to make it - then it becomes out of reach for most people to create a new voice using that new track.

Nevermind - i didnt realize that if you use the "create voice" select from library method it would actually go to the old persona creation system

still it would be easier if it was just in the dropdown

[daring dock]

Hello Suno team 👋

I would like to suggest a few features that could give creators even more creative freedom:
• The ability to choose the exact duration of generated songs and instrumentals
(for example: 30 seconds, 2 minutes, 3 minutes 30, 7 minutes, etc.)
This would be extremely useful for:
intros
trailers
social media content
radio edits
long live or orchestral versions

• A true “Recreate / Alternate Version” feature
The idea would be to completely recreate an existing song with multiple possibilities:
keeping the original vocals while changing the entire musical style
or generating a true full instrumental version of the song
Currently, when creating some instrumental covers, the song is not fully recreated and often feels more like a short extract or partial adaptation.
It would be amazing to generate a complete instrumental version that fully follows the structure of the original song:
intro
verses
choruses
bridge
ending
With the possibility to transform the style into:
orchestral
rock
acoustic
synthwave
jazz
cinematic
80s-inspired
emotional piano etc.

This would unlock huge creative possibilities for:
artists
musicians
singers
instrumental cover creators
karaoke performers
live shows

• The ability to automatically generate karaoke videos
with:
synchronized lyrics
visual effects
animated backgrounds
customizable typography styles
This would be amazing for:
karaoke venues
music communities
TikTok/YouTube creators
live singers

Thank you for all the amazing work ❤️
suno.com⁠� is already revolutionizing music creation, and these features could push creativity even further.

[tepid laurel]

this already exists its just turned off :b well the duration i mean

[daring dock]

So it needs to be activated 😊

[mighty creek]

When it is the playlist update in web, where it does not lag. It lags when you scroll. If you have lots of songs on the playlist. It is unusable in web.

[brittle wedge]

I own the Glastonbury Music Shop and I’ve been using SUNO to write songs about local events and people in our colourful culture. I keep getting asked for CDs by old people! Okay I’m an old person too, I’m 63 but I’ve been making music my whole life and you’re already familiar with a lot of my work. No, you are .. because I hear the songs I worked on when I go shopping .. it’s not an arrogant statement .. it’s just the way it is .. kind of annoying because I’ve heard them all a million times and SUNO is hilariously refreshing and a lot of fun .. a lot of creative fun ..

So I need a couple of things ..

I need a PLAY THIS NEXT or queue so I can serve customers and feel the vibe and drop the next tune ..

I also need the songs to cross fade the last few seconds of a song into the next on the playlist ..

And ideally when I’ve made a playlist for the day, when someone says “can I buy this on CD” I don’t have time to download the tracks one by one I need to be able to bulk download the songs I click on and burn them to CD so I can sell them to customers ..

And while I could prepare these ahead of opening on a bank holiday Saturday when I have tourists everywhere waving money at me ..

It’s about SUNO being the platform that allows me to look at the passing trade .. write a song about the girl dressed as a pink cowboy who’s on a hen do ..

And write a version of “they’re getting married in the morning” and have it on a CD by the time she arrives at the door and says “can I buy this song on CD”

Please don’t ask me why people are buying songs on CD because I don’t understand it myself ..

And PLEASE fix my login because it’s messed up my work flow a lot ..

BTW I have loads of musicians every day joining your platform - I have QR codes on the door and while I also sell instruments I see SUNO as an instrument and production tool that’s worth promoting ..

So ..

Help me out here !!!

[sweet zephyr]

The system, its too harsh or restrictive and lacking context that using a term to describe music is not copyright, IT NEVER HAS BEEN... So IDK who trained the SUNO Models this but it is factually wrong, if SUNO has a legal team, they can confirm this.

[tepid laurel]

i almost never get copyright warnings but its because psytrance is an unwatched genre

[sweet zephyr]

My suggestion is this, well several...

When someone hits the "Thumbs down" don't disappear the thing, allow the user to tell you exactly what is wrong and the model can learn from that and why. Despite the quality of the tunes there has been plenty of really janky low grade generations that just do not fit the muster.

You need a slider that the user can move to indicate the "MINIMUM track length that they are expecting, this is because I am fed up of wasting credits on having tunes that are reasonably expected to be over 5 minutes to be 1:42 in length, like WT???

There is also a problem in terms of a good 90% of what is produced is what is called "Flat" having no other value than its beat content and use would be more background music than Tier 1 stuff. SO... A Variance slider that allows the user to change the amount of expression and variance in a tune like hint would it kill someone to train for key change to assist in the movement of the track?

I shall stop because I got about 5 pages of stuff that I could tell SUNO how to vastly improve its tool, especially releasing a windows stand alone.

[sweet zephyr]

I keep getting one for a track made on here on my account that it refuses to allow as well as a tune I wrote 20 years ago and never released, doesn't like because of copyright...

[west elk]

Yeah. That's how generative AI works. You aren't going to get everything. The instructions are suggestions. It's tries, but sometimes it drops one part to make something else work.

[west elk]

Use the advance settings. There are two sliders to control variance.

[split egret]

People love to use this and pretend that it somehow justifies the issue but it really obviously is nonsense
The problem isn't "it didn't give me exactly what i was hoping for"
The problem is "it asks for direct instructions then ignores them completely and does the opposite"
And it used to be better at following specific instructions that it is now
Meaning that it has been intentionally changed in a way that made it less able to do what it claims to do
"tHaTs HoW aI WoRkS" is just a lazy justification to explain why we shouldn't expect the developers to make their product do what they say it's supposed to do

[dry mural]

Hi everyone 👋 I supposed the following suggestions has already been posted a thousand times but, I would like to ask anyway 😁 🤘

-1st suggestion:
Please, make a million Suno’s user happy with the creation of DAW Suno VST or a something better like a DAW direct interface like the partnership between Splice and Ableton live

-2nd : an AI assistant (web, vst or whatever) trained to analyze a reference source applying specifics tasks usually needed on a pre-mixing stems to master track. The a mixing assistant should have an advanced setting section where fine tuning prompts could be applied

-3rd: last suggestion: please, could devs be able to create the best vat or whatever too for a Setting up or reference trained professional adaptive and ready to stream mastering chain process model?

A partnership between Suno and Ableton live (or any other People fav DAW ) should be one of the most important things in the music production process.

Take care everyone!🔥 (This is a repost, i did the wrong channel past 😅 )

[west elk]

How exactly do you know that "it has been intentionally changed in a way that [makes] it less able to do what it claims?" Are you sure about that? What exactly are you basing that on? Are you an AI programmer? Are you any sort of programmer at all? Because I am. I'm a professional software engineer with a Master's, and have written AI systems. And I'm gonna tell you a little secret: making software system do the exact opposite of what a user wants is actually more difficult than making it do what the user asked. Your whole premise for being upset is ridiculous.

Look. You're upset because AI sometimes doesn't do what you want. I get that. I get frustrated with AI as well; especially Chat GPT.

There are verifiable things these systems should do, but they fail and do incorrectly. But, by design, these systems use millions of input to generate their output, so they aren't simple levers. They don't work the way you think, which makes sense because you clearly know nothing about software engineering or AI development.

So, no, it is not lazy justification of "tHaTs HoW aI WoRkS."

The only laziness is you refusing to understand the tool you are complaining about, which means not just how to use it, but also it's limitations.

Finally, yes, the developers should be working to make the product do what they claim. But the AI will literally never work the way you seem to expect. It may get closer to it, but it'll never get there entirely.

Lastly, in software development, there is a troubleshooting technique where you look at the source of the problem and who experiences it. If everyone experiences a similar problem, then the problem is the system. If only one person has the problem, the problem is the person. (In very simplified terms.)

I've literally never had Suno do the opposite of what I ask. And clearly, most other people don't either, otherwise that's all you'd see mentioned in this channel (or any channel). But it isn't.

Perhaps look at your prompting. Maybe it's overly specific. Maybe your expectations are too specific. Maybe you've given it mutually contradictory requirements and it had to literally just pick one. There are channels here to learn how to improve your prompts. Maybe go ask around in there. I'm not saying the problem is your fault, but I am saying it may be the case that maybe your prompt could be improved to get results closer to what you want.