🧬│ai-chat

[formal jay]

Can I get a voice model of Lina Inverse from The Slayers?

[royal prairie]

Best girl rvc model?
Best browser?

[jovial sleet]

yall got any TTS-sounding voice changers to recommend? like the google translate or TTS bot voice

[jovial sleet]

🥱

[zealous hazel]

you'd have to train your own

[void sail]

yo my mmvc is stuck on booting phase main and not launching the client

[covert lake]

Hello, for help, please elaborate in #1192011222023950368 :)

[clear totem]

-colab

[pine acorn]

📒 Google Colab Notebooks

Google Colab is a Cloud (Remote Good PC) Service. While the Free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

• **Applio**

by IA Hispano
Google Colab

• **RVC Mainline**

by Hina
Google Colab

• **UVR5 NO UI**

by Eddy
Google Colab

• **UVR5 UI**

by Eddy
Google Colab

• **Wokada Deiteris Fork**

by Deiteris & Hina
Google Colab

• **Hina's Modified Original Wokada**

Google Colab

• **RVC-AI-Cover-Maker-WebUI**

by Shiro & Eddy
Google Colab

• **FaceFusion UI**

by Nick088
Google Colab

• **FaceFusion NO UI**

by Nick088
Google Colab

• **Music Source Separation Training (Inference)**

by Jarredou & Makidanye
Google Colab

[elfin spindle]

-rt

[pine acorn]

💻 Local Realtime RVC

Guides for Programs that use RVC Models in Realtime for Calls/Games

• Vonovox

A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE

• Wokada Tg-Develop Fork

A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE

• Wokada Deiteris Fork

Most suggested WebUI with the best general support for many platforms. GUIDE

⚔️ Wokada Deiteris Fork vs Vonovox

For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons

⛔ Outdated/Discouraged

These options are not recommended for use.

• Original Wokada

Not suggested, older versions in youtube tuts are even way worse. GUIDE

• RVC GUI Mainline Realtime

The program is worse compared to the ones above, and much less updated. GUIDE

[random arch]

a

[mortal bough]

I cant seem to figure out why the enunciation of words come out all garbled. If its not that, its that the voice sounds too robotic. What is the best paid realtime voice changer?

[gentle trench]

The good ones are not paid

[gentle trench]

It's most likely the model you're using and what voice changer you're using as well, ask here for help https://discord.com/channels/1159260121998827560/1159290139609137264

[patent osprey]

can anyone else not hear them self

it doesnt say my mic is moving or making noise i uinstalled reinstalled used different voices

and nothing

[digital lava]

gng

wait

theres no more egirl vocies??

[dapper drum]

What are the multiple ways of using AI for dubbing purposes?

[gentle trench]

die

[graceful flame]

Hey , Runway and Invideo AI are a solid option . of course you need go for a paid plans to explore the enhanced features. May be you are trying their free trials that's the reason you didn't figure out more. You can try LunaBloom AI Its a beginner friendly AI Video Generator. No coding , easy to use. It a text to video generator . you can transform your script into high quality videos in minutes. It allows users to create engaging content with AI avatars, voiceovers, and multilingual support. choose the tool as per your goal. if you want blogs or URLs to be converted into videos , you can try out other tools like DeepBrain. Hope this will be helpful .

[still sigil]

how do you always end up replying to messages from months ago

[noble halo]

hello. Im looking to be able to make a voice model that i can make it say what I would like using text to speech. Any tips on how to get started?

I wanna train a voice to sound like it as close as possible

is this possible for free?

sbtk makes rlly good voices and I would like to achieve the same results.

is a 5070ti and 64gb ram good enough for these tasks? rvc locally?

I have tried before but haven't had much look getting good quality results with tts or with training

[fathom sky]

I'm an Entrepreneur and an AI enthusiast. Thinking of creating an AI private online community. Thoughts? too generic? oversaturated?

[gentle trench]

why private? is it sus?

[gentle trench]

idk about the ram but that gpu is def enough for any local rvc task

[fathom sky]

no its not sus, just private for monetization. low ticket

maybe i should make it free first tho

js for conversion rates

[gentle trench]

to get it started off I'd have a free teir and an optional payed thing but not too cheap or too expensive

[fathom sky]

mmmmm yeahhhh

[gentle trench]

don't wanna end up like kits.ai or weights current

[fathom sky]

oooooh yeah true true. Ill start free, especially since paid membership wouldn't make sense if the community has "3 members and 1 member active" 💀

[gentle trench]

yea lol it'd be dead in seconds

[fathom sky]

but what about the concept in general? i think with enough organic marketing it would gain traction

i know it is pretty saturated too

[gentle trench]

I mean it really depends on what you're going for, is it images, videos, voice cloning?

I'd suggest don't do voice cloning as there are free solutions that are just the norm and are really good

[fathom sky]

actually more for info on ai. news, how to use it, research, kinda like this server but more value driven yk?

[fast tundra]

Hey if anyone interested I do sell Instagram pages with 10k 🇺🇸 followers if you wanna grow your model presence

just saying

[fathom sky]

also with more engagement, like weekly calls, gems, etc.

[gentle trench]

ohh I se yea

[fathom sky]

yeahhh

i havent locked in on it yet

but my eyes are set on it

[gentle trench]

tbh starting free with that def is the right approach

[fathom sky]

mmmm yeah i think thats for sure now

I just have to worry about how to build the audience

[gentle trench]

could always start here and other ai servers collecting members

[fathom sky]

oooh yeah, true

[gentle trench]

I'd prob ask around in this chat from time to time to see if anyone is interested in the idea

[mortal bough]

@gentle trench Good morning! I’ve followed the tutorials and tried just about everything, at this point i just wanna know if there are any paid realtime voice changers that are worth looking into

[fathom sky]

yeahhhh, I wonder how the converts are tho, since discord is known to be advertisey sometimes

but worth a try for sure

you guys might hear from me again on here thennn

[gentle trench]

no don't pay, if u really want them to work well just upgrade your gpu on ur pc if you have proformance issues or lag ect

vonovox is the best for nvidia only currently and wokada deiteris is always the best second options and the best for AMD

[mortal bough]

It’s really just the enunciation of words, I’ve bought models from here and still i can’t say words like “going” without it sounding like “gohee” I’m on rtx 3060

[gentle trench]

hmm

ok so first u really shouldn't be buying models u can easily make them yourself for free

ur gpu should run any of the voice changers fine btw

but it could just be outdated models you're using

[mortal bough]

Is because i assumed the model was the issue but i dung think it’s the model

[gentle trench]

what voice changer are u using anyways currently?

[mortal bough]

I’m using vonovox currently most updated version

[gentle trench]

good luck Icey

[gentle trench]

that's odd then, maybe the settings of vonovox itself? I doubt it tho

[mortal bough]

Do you use index files?

[gentle trench]

what..

oh

I use index but I always keep it at 0

there's never a reason for me to change the usage of it as it's best at 0

[mortal bough]

Ah ok. I find no difference. I also thought that could be why the voices couldn’t pronounce certain words

[gentle trench]

Have you talked to Dr87 about the issue btw? he's the creator of the app so I'd go to him personally to see what he thinks

[mortal bough]

I will try that. Thanks!

[gentle trench]

np!

[mortal bough]

i always make sure to run audiodg on high priority cpu 2

[gentle trench]

what is audiodg?

[glad nebula]

bad pronunciation in realtime can be either the model or the microphone not being clean enough

since the audio gets resampled to 16k, the source has to be good quality

[mortal bough]

Noted. Could you recommend a model so i can rule that out?

Could be my mic but i thought i had a decent one

[glad nebula]

idk i dont use other people models

[gentle trench]

shh give her one of yours

[glad nebula]

[mortal bough]

FIFINE Studio Condenser USB Microphone Computer PC Microphone Kit with Adjustable Boom Arm Stand Shock Mount for Instruments Voice Overs Recording Podcasting YouTube Vocal Gaming Streaming-T669

I think i tried and failed making a voice model before

[glad nebula]

as long the input audio doesn't have loud background noise, should be fine
i'd recommend trying a different model and check if the problem persists

[mortal bough]

I've tried many models, that's why i was asking if you could just recommend one you know is good quality so I can rule it out. Idk how to verify a model is good quality

[glad nebula]

all of them had the same pronunciation problem? hmm if thats the case then is the mic

[mortal bough]

They sound fine when i use an input file but real time, not so much

[glad nebula]

when you're doing local conversion the model sees the whole audio at once, thats why it produces better result

[mortal bough]

I guess spending $50 on a mic is considered pretty cheap. Those things get pricey

[glad nebula]

hmm are you too far away from the mic?

[mortal bough]

No, and i use that filter screen and keep the volume or gain on it low

But background noise is an issue

Cuz I live in nyc

[gentle trench]

my mic that I use is the one that comes with my valve index vr headset which is pretty good for realtime but I'd think any somewhat decent gaming headphones would have a good result'

[mortal bough]

I've tried third party apps for that issue

[glad nebula]

have you tried vonovox? it has realtime denoising

[mortal bough]

I have vonovox

[glad nebula]

i have a 50$ fifine mic too

try enabling rnnoise

[mortal bough]

Well thanks, i think you guys are helping me narrow it down to being a mic issue

[glad nebula]

yea i suspect is the mic

[zealous hazel]

looking for an rvc model with like a futuristic TTS female vibe to it, dystopian anybody know something good

[edgy light]

No eGirls allowed

[zealous hazel]

its for an edit im making of the new sora 2 and meta vibes announcement

https://www.youtube.com/watch?v=kBkKxi69gGQ

https://www.youtube.com/watch?v=dNrTrx42DGQ&t=678s

[dusk mirage]

Guys, can i ask if there is an AI that can make music?

[zealous hazel]

suno

https://suno.com

[dusk mirage]

Hmmmmm. Is it good?

[zealous hazel]

its the best

[dusk mirage]

Let me try

[gentle trench]

I know one yea, search for GLaDOS

[zealous hazel]

not what im looking for, currently trying out to see if i can get what i want with the default 11labs plan

[dusk mirage]

Btw thank you for your suggestion

[molten juniper]

how i do my own ai voice model ?

[zealous hazel]

upload it to youtube

[thorny helm]

-rt

[pine acorn]

💻 Local Realtime RVC

Guides for Programs that use RVC Models in Realtime for Calls/Games

• Vonovox

A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE

• Wokada Tg-Develop Fork

A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE

• Wokada Deiteris Fork

Most suggested WebUI with the best general support for many platforms. GUIDE

⚔️ Wokada Deiteris Fork vs Vonovox

For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons

⛔ Outdated/Discouraged

These options are not recommended for use.

• Original Wokada

Not suggested, older versions in youtube tuts are even way worse. GUIDE

• RVC GUI Mainline Realtime

The program is worse compared to the ones above, and much less updated. GUIDE

[noble halo]

Ram isn’t enough 😭

[gentle trench]

💔

Give your wallet to the ram gods

[noble halo]

How is 64 not enough 😭

[gentle trench]

I really don't know

[noble halo]

Welp. Ig I ain’t doing it then 💀

[light current]

Could sombody help me fix the voice changer

Please?

[gentle trench]

https://discord.com/channels/1159260121998827560/1159290139609137264

In here yea

[static swift]

@obtuse burrow search peterbott

[lusty onyx]

hello anyone recommend for me course about ai please

as a begaineer with rich and solid start point

[gray flare]

Dude to get honest I learn to use the most of the chatbot learning with own gemini 2.5 pro, But do you want to learn more than just using the right chatbot?

[dusk forge]

can someone pls send me their sora invite code

[grave lodge]

How do I recreate Cartis voice without using ai on bandlab

[gentle trench]

Autotune

[void aspen]

fweh

[gentle trench]

Fweh

[manic thorn]

Hey folks, do you think it'd be possible to make a Loab ai agent built for pen testing and brute force applications?

I've been talking to my chat bot about it, I'm kinda into horror and things like that so I like the type of allure this topic brings.

[warm cosmos]

how to download?

[covert lake]

I asked you to elaborate in #1192011222023950368 so i can help you, this is a general ai server

[covert lake]

autotune and training your own voice

[covert lake]

that’s a random user, not a bot, You can search rvc ai voice models at:

• https://discord.com/channels/1159260121998827560/1175430844685484042
• In https://discord.com/channels/1159260121998827560/1163592055830880266 , Do /find with @hidden grotto
• https://weights.com/ (login required)
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• https://voice-models.com/ (just scrapes the huggingface link from the voice-models channel)
• https://thevoicemodels.com/ (for Turkish Models, login required with discord and level 2 on their server)

if there isnt one, you can:

• make it yourself with our docs guides
• Ask a free request in https://discord.com/channels/1159260121998827560/1159289738314919936
• Be aware that we don't allow any paid comms in the server

[hidden grotto]

:wave: @covert lake, How can I help?

Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models

[covert lake]

yeah

[graceful flame]

Hey , I'm new to this platform so i used to scroll all the feeds even the older one so if any information i get , or anything i can share that i knew well, that's the reason you can that im replying for old questions also.

[sage moss]

Does anyone have a invite code for sora 2

[still sigil]

theres no way you scrolled back 3 months just for that message

[dawn temple]

I think you can find them on Ebay

[sage moss]

I'm not paying for one 😭

[floral aurora]

Hey everyone, I’m still kind of new to AI Image Generation so I wanted to ask something.

I’m trying to keep my character consistent across different prompts. I uploaded my character’s image, and then in the next prompt I gave it another reference image (different pose, different angle) and asked it to generate my character in the same style. But instead of giving me a new pose, it just gave me the exact same image again.

Is there a way to make Nano Banana keep the same character design/body but actually generate new results in different poses? Can I do this directly in Nano Banana with Gemini, or do I need to try another tool for better character consistency? Any recommendations would be super helpful.

[zealous hazel]

hes lying there are no 10k instagram babes, i sent the bitcoin and then got ghosted please report this guy

see?

[hasty ruin]

hi

any here

[full ore]

Do you think they'll fix the long queues or is it permanent? Also I'm waiting until I have a stronger computer to run Replay.

[covert lake]

It might be there are many people using Weights.com at the moment, you could wait or try premium for shorter queues iirc

It's better to ask in discord.gg/weights

[full ore]

You'd think people would leave after the paywall update. Speaking of updates, I have a feeling Weights is about to have a massive "pay-to-do-anything" update hence the long queues.

[covert lake]

@dawn temple are you aware that Weights lost its vanity? it only has 11 boosts currently

[twilit bluff]

Hello I am a new

[full ore]

I wonder how many people got banned from the Weights server because they hated the First Paywall Update, with a subscription only update on the way. Also the invite is invalid.

[gentle trench]

You poor soul

[gentle trench]

Weights died a long time ago

They're just burying themselves in a late grave

[full ore]

When I had my accounts things were fair: You could get free tools to train models, download vocals and instrumentals, and could do 100 covers a day. They could've done something like Patreon donations, and offer rewards better than plans

[gentle trench]

If only

[full ore]

I even have a meme about what they would do, and if I was on the Weights server I would show it

[gentle trench]

They'd probably make a paid subscription to breathe if they could

And say it's to cover the cost of the gpus

[neon spire]

hi

[elder willow]

hi i got question whuch model best for e girl

[elder willow]

hey

[gentle trench]

Have you ever read the rules

Freak

/j

[full ore]

Someone should spread my beliefs

On how Weights can generate money and bring back the Old Limits. It would actually save them from having to potentially import their "Paid Subscription Only Update"

[still sigil]

sounds like something a cult leader would say

[full ore]

I'm not a cult leader, I'm one of the various angry Weights users who left the site after you-know-what.

[safe storm]

do i download okada voice changer or i try smt else

bc idk whats the best tbh

[lime sleet]

best settings for female to male voice?

[safe storm]

ur trans ? just asking ❣️

[dawn temple]

Ohh, I knew that we lost boosts, but didn't know we lost vanity as well 😆

[wet wedge]

who’s @neat sail

[dawn temple]

Abby

Why are you in ai chat Kia

[wet wedge]

who’s abby

[dawn temple]

🫢

[wet wedge]

OH

js realized

[dawn temple]

Hmm, you remember tic and tac from Dreamtavern?

[wet wedge]

no..

[dawn temple]

🥀

🌾

[wet wedge]

STOP

[dawn temple]

hahaha

[gentle trench]

I know what you are...

[polar flax]

bro want to spread misinformation

[abstract totem]

Weights is here in Gotham

[lime sleet]

No I'm on my friend's discord account mine got suspended and I wanna troll someone 😭😭

[lime sleet]

What am I...

[pearl socket]

can anyone help me with finding a convincing female voice?
i also need its settings

[polar flax]

I don't think we'd support you for that intention

[solemn mural]

guys, how do i delete the index file in rvc?

[solemn mural]

just delete in the explorer?

okkkk

[chilly lake]

well

if you mean realtime voice changer

then just delete the model on UI and re-add it without index

or just set index use to 0

[solemn mural]

i deleted it, recompiled it and i got "unkwon" error

but

index file and .pth file deletted

soo

thx i guess

[lime sleet]

Uh ok

[upper tendon]

Do you guys like AI

or is that not one of the topics here

like ai ethics

[queen stone]

-rvc

[pine acorn]

📚 RVC Documentations

AI HUB Docs

https://docs.aihub.gg

🍏 Applio Docs

https://docs.applio.org

[covert lake]

@dawn temple https://www.youtube.com/watch?v=f9HwA5IR-sg this is crazy, all sources are in the description

it's an anthropic experiment

[dawn temple]

🫢

I think I have heard something similar

I also watched podcast of godfather of AI

[rapid blade]

hi

[odd knot]

h

i

do y'all think i still take the crown for the best sounding models when it comes to vtuber models and anime models?

despite some of the models i made are scuffed

[gentle trench]

You're going to be castrated if u continue ❤️

[pearl socket]

Huh?

[gentle trench]

you know what you did

if u read the rules

[solar torrent]

[daring sand]

Hello

Give me code for sora pls

[gentle trench]

https://cdn.discordapp.com/attachments/1159290139609137264/1419738795833819269/Jaxcostume.gif

[pearl socket]

I just read them, which did i violate?

[gentle trench]

egirl one dum dum

now don't do it again or ur weenie is getting snipped off

✂️

[indigo wind]

Is there any way of combining w-okada with voicemod so as to have the voice from the former and the soundboard of the latter at the same time

I want the voice from wokada but the soundboard from voicemod and be able to use both at once

[gentle trench]

pretty easy actually yea, just have to use vb cable over vac lite

[indigo wind]

I got vb cable what do I do

I tried effing with the inputs and outputs on voicemod and it doesn’t like that and says audio disconnected, I.e I set both input and output on voicemod to vb cable and got some error

[gentle trench]

input on voicemod should be cable input and output is just whatever headphones u use to hear, and on the voice changer use just your regular headset/headphone mic and output should be cable input

[indigo wind]

This works with shit like VRChat?

[gentle trench]

yup

[indigo wind]

Alr ill try that tonight thank u

[gentle trench]

if u want it to sound better tho with free fx on fl studio I could send u the setup I use, you'd just need to download vac lite, fl studio, and voice meeter

[indigo wind]

I think I’ll stick to this for now but thank you for the offer

Should the one in VRChat itself be voicemod or vb cable

[gentle trench]

oh yea in vrchat keep ur mic as voicemod

[indigo wind]

Oh yeah maybe that’s why it didn’t work I had it set to vb cable

[gentle trench]

lol

[gentle trench]

did it work btw?

[indigo wind]

I haven’t got on yet to check but I will tonight

[gentle trench]

could add me on vrc so I can check just u and me so u don't sound silly in front of a bunch of randoms

[indigo wind]

Sure

[gentle trench]

https://tenor.com/view/cyn-cynessa-murder-drones-md-cyn-frog-blink-gif-16166011548669509749

[indigo wind]

What’s your name on dere

Woah my name is colorful wtf

[icy frost]

I am fedya

Fedor

[indigo wind]

hello human

[icy frost]

Are you not a human?

[indigo wind]

No am eldritch program that destroyed the earth

[icy frost]

I love it

[icy frost]

I am fedya

Just remember

You will hear this name again.

I am a legend

You know ?

[indigo wind]

Fedja

[icy frost]

Fedya

I am from russia

[indigo wind]

HAHA

https://tenor.com/view/fnaf-deleted-i-saw-what-you-deleted-gif-7532358250017057212

[indigo wind]

How do I fix w-okada misinterpreting what I’m saying

Like I’ll go to say fuck and it’ll say suck instead

[gentle trench]

what version of w-okada are u using?

btw should be askin in help channel

[indigo wind]

Oh alr

[indigo wind]

Uhhh lemme check

[indigo wind]

vcclient_win_cuda_2.0.78-beta.zip

Most words it works but it occasionally slips up and does the wrong letter

Idk if it’s the ai model or just how w-okada works

[gentle trench]

Oh yea you need some new stuff that's olddd

It's collecting dust

What gpu do u have if u told me before I forgot lol

[barren pivot]

hello anyone can sahre a good people detection model?

[indigo wind]

I don’t have it memorized

Where do I download the latest w-okada?

[gentle trench]

Check your performance tab on task manager

It matters to which one u have, either Nvidia or AMD

[indigo wind]

I know where to find it I just can’t get to my computer rn

How come the w-okada version I have is out of date I got it directly from their site

[coarse flame]

sora 2 code please

[indigo wind]

How do I get the newest version?

[gentle trench]

Just newer better ones have been made, modified versions of the one u have

[gentle trench]

I messaged you in the AI help channel, if u go to the guide for each one it tells u where and how to download and use it

[indigo wind]

Thank you!

[gentle trench]

Np! If u need more help just ask me here or dm me

[indigo wind]

Will do!

[arctic sequoia]

sora 2 code please

[gentle trench]

no

go away

can we get a bot that times out anyone saying this shit for 3000000 hours

[limber nexus]

i could make one, just gotta get whoevers in charge to add it to the server

[gentle trench]

it was a joke but the ppl asking for the code is kinda annoying lol

they're probably making new accounts to come here from who knows where just to ask for a code to sora for some reason

[limber nexus]

i know, i just thought it would be funny to have a bot that would ban more bots

[gentle trench]

that would be a good idea actually

it should only ban/kick/timeout the person if their account was made this year tho bc they obv just made it to be annoying

don't want it accidentally kicking someone who matters

[limber nexus]

and it could have a AI that would track messages to figure out what the spam messages are like "sora 2 code please"

[limber nexus]

right, forgot about that

[gentle trench]

yea!

[limber nexus]

but the hardest part would be finding hosting for the bot with the AI

[gentle trench]

oh true yea

[tired yew]

hi i was wondering if anyone here is available for voice model creation for commision?

[gentle trench]

it's actually pretty easy to learn yourself, it's free and u can do it via locally on your own pc or on a website which is what I use

[crude yarrow]

hi yow

[gentle trench]

https://tenor.com/view/wave-emoji-smile-wave-smiling-waving-emoji-haii-gif-7959794316942733877

[polar flax]

only free requests are allowed https://discord.com/channels/1159260121998827560/1159289738314919936

[umbral bloom]

yo is anyone using sora ai for some odd reason mines not working so well right now

[gentle trench]

I swear

[umbral bloom]

u too?

bro it keeps saying third party content its so annoying

[gentle trench]

there's been like a billion (3 people) who have been asking about some dumb sora code pls be different

[umbral bloom]

What?

[urban token]

[mental vapor]

#боюсьженщин

[pulsar pivot]

yo what ai client do u guys use?

[covert lake]

Well there are many different AI models, types and programs

[covert lake]

@tired yew Paid Commissions aren't allowed

[crimson zodiac]

Hi guys, does anyone know ai which can create 30mins plus songs in lofi style?

[crimson zodiac]

Im using braini ai on phone

[fair whale]

where can i get voices?

[covert lake]

you mean RVC Speech-To-Speech voice models? #1175430844685484042 and https://weights.com/models

[gentle trench]

What

[covert lake]

@dawn temple I looked into some old bookmarks I had, and damn many AI sites shutdown

[gentle trench]

what kinds?

[covert lake]

any

podcast creators

[gentle trench]

any notable ones?

[covert lake]

voice

[covert lake]

idk those were some old bookmarks

the last i just cleaned up was podchat

[gentle trench]

how old we talkin? like 2023?

maybe older?

[covert lake]

AI boom, so 2023-2024

bro got acquired

@gentle trench old 2023 ai hub rvc gem? https://rentry.co/nh2arh

[gentle trench]

yoo it quotes our server

[covert lake]

meta voice

[gentle trench]

I blame ngrok

[thorny storm]

does anyone know how i can find the file [ start:http] something that opens the ai thing where i do the voice changers beacause i lost it

[gentle trench]

how do u lose the file that starts the voice changer

:(

[thorny storm]

it was something that opened something like a cmd panel with a code and after a few seconds it opened the voice changer app

i accidentally deleted it

[gentle trench]

why 😭

what version of wokada are u using

or is it vonovox

[thorny storm]

idk abt what ur saying i found this on a catching pedos video in yt

[gentle trench]

then u def have a really old version

outdated and crusty software

what's ur gpu

[thorny storm]

rtx 3060 12gig

[thorny storm]

i got it 6 plus months ago so

[gentle trench]

def super old and outdated

[gentle trench]

use vonovox it's the newest best one out there, there's a guide bt all u need to know is just download it and run the setup.bat then start.bat

[thorny storm]

do u have a yt vid tutorial?

[covert lake]

all youtube/video tutorials for ai realtime voice changers are outdated

check https://docs.aihub.gg, and ask for help in #1192011222023950368 following the help guidelines

[gentle trench]

have you heard of this super cool thing called reading

just open the setup.bat then once that is done run start.bat

not hard

[umbral bloom]

does anyone know how to bypass sora ai copyright

[gentle trench]

hell no

https://tenor.com/view/annoying-orange-annoying-orange-rage-rage-phone-destroyed-annoying-orange-phone-gif-9984038394419705078

[gentle trench]

nah

[limber nexus]

https://tenor.com/view/brokengjh-gif-20992403

SO MUCH SORA 2 STUFF

[leaden iris]

If u find a way dont forget to tell me

[umbral bloom]

bro im struggling to find one

[leaden iris]

Me too

[umbral bloom]

even trying to use non copyright charecters they still flag it as copyright

[leaden iris]

The world is too cruel for us

[umbral bloom]

u cant blame them bro jjk and chainsaw man animators were crying when they saw the yuji vs denji fight

that was so sick

[leaden iris]

Thats called Freedom 🇺🇸

Btw.. why they are making so delay with making s3 of jjk

[knotty sequoia]

chat whats the best ewoman voice model

[gentle trench]

you're joking

[knotty sequoia]

?

[gentle trench]

go read the rules again

[wanton pilot]

Probably a dumb question but I'd like to know given yalls experience with ai voices.whats the overall benefit of using these compared to a paid service (voicemod) other than the obvious (community,voices,free)

[gentle trench]

what gpu do u have?

[wanton pilot]

Nvdia GTX 1060 off the top of my head (could be off on the last two numbers lol)

[gentle trench]

the benefit from my experience is not having to pay a software money to make my voice sound different when playing as a character like Venom or Darth Vader, voicemod could never even make those voices possible for someone like me who cannot get a gravely voice or some of the voice effects used on official stuff being payed which makes voice models useful in that way for ai voice changers

[dapper drum]

What are the multiple ways to dub something using AI?

[still sigil]

youve got to stop spamming this

[dapper drum]

😔

[gentle trench]

They're a bot/j

[thorny storm]

i cant find the setup.bat in the folder

[prisma cradle]

is there a way to like stop delaying

and make it smoother

i dont know how to say it tbh

[covert lake]

For help, read the help guidelines and elaborate in #1192011222023950368

[gentle trench]

That cat looks like vidal

[gentle trench]

[vernal roost]

Hello - new here

[dapper sparrow]

GUYS what is a good LM model i can run with my rtx 3080 fe? i want something good all round. chatgpt told me mistral 7b is good.

[chilly lake]

10GB or 12GB?

[covert lake]

nop

[dapper sparrow]

i think 10gb vram, i have 16gb ram

[chilly lake]

if you download lm studio there are options to choose which quantized model to download based on your vram size

8b is probably the best choice for the parameters

the model itself depends on what you wanna do with it

16GB RAM is also very limiting

you got a 5-year old gaming PC, not an AI box

[dapper sparrow]

😭

the rtx 3080 is still my baby

one sec

overall smart lm to talk with debate with and reason. i want it unbaised as possible if that even exists. but mainly intellegent. im not using it for coding or developing anything

also wnat to use it to organise stuff

chat gpt told me this one is good: Mistral 7B / the “Mistral-7B Instruct v0.2 (GGUF)”

then someone else recomended me gpt-oss-20B but chatgpt said it hallucinates alot and mentioned some other problems, also due to my specs might not be the best for me.

[chilly lake]

20b is too big for you

[warped steeple]

Is there a 50 Cent vocal sample?

[paper wigeon]

How would I run a chatgpt code interpreter like llm locally, im tired of getting cucked by the 10 messages per 5 hours limit

[steep night]

hey

[distant rose]

?

[jaunty tapir]

Hello

[polar tundra]

Any modern AI model comes with built-in restrictions, but it's possible to mitigate some of these limitations using designed prompt engineering techniques. For local deployment on a system like the RTX 3080, I'd recommend either a quantized version of NVIDIA Nemotron-Nano v2 or one of Qwen3-4B/8B (thinking). These models are efficient for conversational and organizational tasks. There are also open-source repo on GitHub designed to optimize controllability and reduce bias, or you can easily make one yourself..

[polar tundra]

Open WebUI if you need a ChatGPT-like interface is a good choice. If you're using it for programming, Kilo Code and Qwen Code CLI are stronger and more appropriate.

My ranking for programming:

1. Qwen Code CLI
2. Kilo Code
3. Open WebUI

There are installation instructions for Qwen Code CLI and Open WebUI on their respective GitHub repos. For Kilo Code, it's a VS Code extension available on the marketplace.

[lament arrow]

So it looks like Sora 2 always had the ability to not do copyrighted works, but as usual move fast and break stuff

only thing I'm concerned with now is when can the non-commercial versions catch up

[indigo wind]

do you guys know how i can hear my self on vonovox

[frozen egret]

hey whats the best place to make ai covers? most i can find on websites now require for me to pay a subscription

[gentle trench]

weights.gg

[frozen egret]

that's the thing tho, i use it too and it also uses subscription

[frozen egret]

what about open source ones

i heard it's better than weights.gg

[still sigil]

weights replay

[frozen egret]

whats that

[gentle trench]

applio is free, if u got a good pc for local

[frozen egret]

no i use laptop

what about sumn google colab shi

[frozen egret]

@still sigil whats that cuh

[still sigil]

you make the cover on your laptop

https://www.weights.com/replay

[frozen egret]

is it a free version of weights?

[still sigil]

its free but it uses your laptops resources instead of weights server

[frozen egret]

3GB is not that much

[still sigil]

it does covers and training voice models if you have a 30 series + gpu

*covers dont need a gpu

[gentle trench]

only thing is u gotta get all the models by downloading on ur pc which is annoying

even tho it should fr load your models if u made any using email or discord acc

[still sigil]

i mean how would it load the model if it doesnt download

since its running the model locally

[gentle trench]

By finding it off weights database

[still sigil]

thats still not running locally

[gentle trench]

It shouldn't need an account then to use it

Kinda pointless for the local program

[minor ravine]

Hi

[dapper sparrow]

Yh that's what I got qwen 3 4b thanks

[deep wharf]

hlo

[coarse blade]

What is the best voice changer ?

[gritty gulch]

are there any girl headset models

[polar flax]

try find them in temu

[gritty gulch]

dude

[polar flax]

you said "girl headset", perhaps the fancy pink ones or with cat ear accessory?

[gritty gulch]

i mean voice models

like that sound bad

[spark agate]

Have there been any rvc or ai updates lately?

[gentle trench]

E-girl trolling huh? Ever heard of not breaking the law

[prime loom]

My only question is why…

[polar tundra]

No problem

[prime loom]

Idk but I feel like ChatGPT became more “emotional” like gpt 4 but also now has reasoning like gpt 5 (I have no clue don’t quote me on this)

[brazen juniper]

Badges Progress

Collector (Upkeep Required)
-# Earned for completing all bundles. The most expensive badge to achieve.
0% 0 / 28
-# Earned for completing all regular achievements and non-upkeep gold badges.
0% 0 / 96

Daily Devotee (Upkeep Required)
-# Earned for a "high" daily streak.
0% 0 / 420
0% 0 / 666

Poor Little Beggar
-# Earned for successfully begging.
0% 0 / 10K
0% 0 / 15K

Hide and Seek Champion
-# Earned for successfully searching.
0% 0 / 10K
0% 0 / 15K

Crime Lord
-# Earned for successfully committing a crime.
0% 0 / 10K
0% 0 / 15K

Meme Connoisseur
-# Earned for posting trending memes.
0% 0 / 2K
0% 0 / 3K

Page 1 of 6

[wheat ridge]

帮我写一篇关于人工智能发展的100字文章
或者
总结一下今天的AI新闻

[gentle trench]

I'm scared

[drowsy heath]

Hello

[gentle trench]

I do not trust that

[dapper drum]

Has anyone ever tried using synthesizer V for backing vocals?

[gentle trench]

that's for vocaloid models

like Miku or Teto

[hasty zinc]

Fool

[gentle trench]

I trust that even less

https://tenor.com/view/oh-moderators-gif-25571015

[warm slate]

Hey, does anyone know any good voice ai? Something that is text to speech but not as crappy sounding

[crude nacelle]

Yo

there's some troll in the voice chat

being racist

and doing very inappropriate shit

their user name is X1wend

[gentle trench]

what are they doing?

[glacial pike]

VIX @vix._.envix are you OK? People just want ask about your voice model, how to set it up correctly. Why would you them unfriend and block them? They would message you, adding you as a friend. Because Discord, "This user only accepts messages from friends."

[gentle trench]

who are you @ing ?

oh

yeah he went through something. leave him alone please

I don't have contact with him, he's not in this server at all

[covert lake]

Yeah that was a scammer that got banned

[glacial pike]

Then delete his vocie model from this server

[gentle trench]

no? he has good models we don't need to remove good stuff so more slop can get put in its place

[glacial pike]

But people have questions for the author, you understand that, don’t you?

[gentle trench]

I do yes, but u can't ask him if he's not here

try asking him on weights

[glacial pike]

You see, the problem is that he accepts friend requests and then deletes and blocks them. He doesn't even want to answer questions, but as the author of his voice model, he should at least answer one or two questions, right?

[rugged quarry]

67

[gentle trench]

it's up to the person, I wouldn't mind answering questions about my models but if he doesn't want to talk it's his choice tbh

[glacial pike]

Of course, you have strange concepts, but God will judge you.

[gentle trench]

weirdo

[glacial pike]

I'm Harley Quinn and yes, i am 🤪

[gentle trench]

that is the most normal thing u said so far

[glacial pike]

Well, if you're not my fan, we're not on the same page, buddy. 🙂

[gentle trench]

I prefer Mr J tbh

[glacial pike]

Pudding, lol

[stoic warren]

hey guys any ressource to use my own skill in blender and animation to drive AI generation result ? i have asked chat gpt im just curious if anybody has define method or know some

[limber nexus]

and answering a question was never in the 10 commandment's last i checked

[glacial pike]

https://tenor.com/view/facepalm-cookie-monster-gif-14188411006779371679

[limber nexus]

everyone has a right to privacy, even if its something they posted, and as ive already seen said he went thru some stuff, so im leaving this conversation now

[stoic warren]

im i at the wrong place to ask that or is that just cause peopel dont care or can't answer ?

[crude nacelle]

Saying the n word, and moaning

[gentle trench]

ew..

[crude nacelle]

Yeah

[gentle trench]

that's loser behavior I hope that person burns in broad daylight and never gets to see heaven

no excuse for such words

[crude nacelle]

Yeah I don't like racism

or trolls

that's a super bad combination

[gentle trench]

trolls are ok but racism is not at all

[crude nacelle]

Yeah fair

[gentle trench]

it depends on the kinda troll

if they're just being weird it's tolarable

[crude nacelle]

I had em muted

[gentle trench]

perfect

[crude nacelle]

sounded like he was using some kind of simulated sound board

[gentle trench]

bro can't even use his real voice

[heady briar]

hey guys, does anyone have a sora code they could spare?

[gentle trench]

we don't have any sora codes here shoo

[heady briar]

damn alright im sorry

[steady surge]

hello fuckers

zamn rank gong

how do use rvc

[steady surge]

what fork do people use nowadays

[gentle trench]

for training or for voice changer

[steady surge]

both

[gentle trench]

for training I use Applio which is waht most use and for voice changer it depends on your gpu

[steady surge]

do people normally still get their models here or do they just get them from weights now

[gentle trench]

if u want good stuff get models here

[steady surge]

is weights still just "we steal all our models from the discord server without user permission"

[steady surge]

do you happen to know what the best vocal separator is in uvr now?

last i used it i was using roformer instvoc duality

and does applio come with pre trained models for training

or whatever you call them

base models

[gentle trench]

yes

[gentle trench]

for making datasets I use gabox's vocal fv4

then use reverb removal by anvuew and de-echo normal

and if dereverb destroys the voice of the character like if they have fancy effects on their voice and it removes those or the voice entirely use the de-echo de-reverb

[steady surge]

do i need to split audio clips into individual files if they are all a couple seconds or should i just leave them all in one

[gentle trench]

Just leave the dataset in one long file as applio slices it for u when training

[steady surge]

how do i select a pretrained model?

i downloaded titan from the download menu

[steady surge]

was the default

but idk how to select it in training

[gentle trench]

Oh don't use that one it's an outdated bad pretrain, the only good pretrains is the original, and legacy core V2, the only version of V2 is the spinv2 version for legacy core

Original is for contentvec

[steady surge]

i dont see original in the list

[gentle trench]

It's used if no pretrain is selected

[steady surge]

oh

[jaunty tapir]

Has any1 used sora yet

[gentle trench]

If you're asking for a code get outta here pretty please

[jaunty tapir]

Didn’t

Was gonna tho

[gentle trench]

Yea, if u wish to use the original pretrain deselect the pretrained button

I'm applio

[steady surge]

if im training 1k epochs, can i just save every 100

[steady surge]

i dont even see a pretrained button

[jaunty tapir]

Ngl I’ll pay $5 for a code

Hold up someone else reacted rose same time I did

Nah what are the chances

[gentle trench]

You should be saving every 5 epochs, and train up to 900 at most, you will not need it to go that high but most models are good after 100 and before 600 usually

Gotta listen to them and check the tensorboard

[steady surge]

what if im lazy

[gentle trench]

You won't know which one is good then

[steady surge]

the input is shit so any output is going to be equally as shit

so idc

[gentle trench]

You don't have to check every single one but like the good ones, the line that goes up is bad and the line the goes down is good

[steady surge]

idk how to use tensorboard

so

[gentle trench]

So check the low spikes, that's what I do but trust your ears over what u see

[steady surge]

is it just the batch file

[gentle trench]

What batch file?

[steady surge]

run-tensorboard.bat

[gentle trench]

You mean batch size? That depends on how long the audio is

[steady surge]

.bat is a batch file

[gentle trench]

Oh wait are u using it locally or on a web browser like kaggle or Google colab

[steady surge]

local

[gentle trench]

Oh

Idk how to use it locally but I know how to use the interface of applio and tensorboard

[steady surge]

im tensoring

i have no idea what the fuck any of this shit means

id send a screenshot but the mods chastised me for leaving the server and then rejoining

[gentle trench]

You gotta get your roles back i guess then

[gentle trench]

Why'd you do that btw

[steady surge]

because this server annoys me

but also i needed to train and forgot how

so i rejoined

[gentle trench]

How? Some of the people in the help chat are pretty dum sometimes on how they ask questions like "how use rvc" and stuff

I laugh at it

[steady surge]

because weights took it over

was more fun before that

now the server is an advertisement

[gentle trench]

Weights sucks but it's trying to save itself, I'm letting it die

[steady surge]

i came from the old server

before it was deleted

i used to have one of the highest viewed ai cover youtube channels before youtube started nuking ai covers

[gentle trench]

Copyright is a bitch, killed the old server which is why all samples posted for models are just vocals

[steady surge]

https://ibb.co/3yktNkZV what does it all mean

[steady surge]

images wont even embed

so cucked

[gentle trench]

I can still view the link it takes me to the image dw

[steady surge]

that image is at epoch 300 i think

i have no idea what it means

congrats Mr bot

you did it

[gentle trench]

https://tenor.com/view/oh-moderators-gif-25571015

[gentle trench]

basically the line that goes up is bad line that goes down is good

[steady surge]

aight

your business can suck my business

waist line

[sharp glacier]

does anyone know why it says dml

[lyric cove]

how do i get a code for sora 2

[hazy vessel]

anyone got a soda code for me??

[polar flax]

candy crush soda?

[heavy carbon]

Yall i need help what is better RVC OR MMVC?

[red cape]

HI?

[heavy carbon]

I have a good cpu and gpu

[limber nexus]

Hi

[polar flax]

@quaint iron please go to https://discord.com/channels/1159260121998827560/1359898289335566570 if you want to share an open source project (any kind of proprietary/commercial ones are not allowed)

[gentle trench]

those are two completely different things

[noble halo]

Hello. How do I make ai voices like sbtk for free? Is 64gb ram not enough?! 😭 I have a 5070ti 16gb vram

They make good realistic ai voices and it sounds just like the characters?

I would rather not pay for tools 😭

I need some more simple in depth info.

Probably rvc.

If someone could dm me and walk me through it that would be amazing 😭

I want it to basically make the voice the same but use text to speech to get it to say what I want?

[limber nexus]

ok, i can't help with any of the voice model stuff, but i think you would be able to generate the TTS then kinda put the voice model over it to make it the voice you want

[polar flax]

if it takes more resources than your ram+vram, you must have used improper rvc training configurations, such as batch size more than 8 for fp32 precision

or if it is another voice cloning solution that may be too demanding for your spec, you should stay away from it

[marsh lagoon]

I've tried building a model using 'legacy core v2', and I felt that its Japanese performance is a bit lacking. I guess this might be because I couldn't use Japanese Hubert

[gentle trench]

use applio to train models, i'd recommend original pretrain or use the new legacy core v2, v2 for content vec will be available in an hour or 2

[noble halo]

Vec?

[noble halo]

Idk what pretrain or legacy code is. Let alone applio to train 😭

[glad nebula]

spin v2 was trained in english, yeah

[leaden rain]

@south eagle

[glad nebula]

the pretrain itself was trained in chinese, but that doesn't matter as the model forgets the pretrain language by epoch 1

japanese hubert, chinese hubert etc needs pretrains tho (or maybe not, uhh did they work fine with the og pretrain?)

[polar flax]

pretrains using jp hubert are actually bad

[marsh lagoon]

Personally, KLM4 worked well when I used it with Japanese Hubert

I could barely notice a difference in timbre, but there was a big difference in the Japanese pronunciation

[glad nebula]

i think they work fine

idk i havent never tried them

as long it's a contentvec pretrain it should work i guess

[marsh lagoon]

Also, legacy core v2 seems slightly more stable

[glad nebula]

ive got the contentvec version of legacy v2 now

im gonna upload it on a few minutes

(voice of the pretrain)

[marsh lagoon]

In my opinion, while there's barely any noticeable difference between using jpn hubert and not in sections where the inferred vocal's pronunciation is clear, a big difference ultimately emerged in the parts that were faster or heavily mumbled

[glad nebula]

yeah, sadly i cannot train legacy v2 with the japanese hubert as the dataset is chinese

contentvec/hubert_base and spinv2 only knows english

ive heard they also struggle in fast phased speech/singing, even in english

which i think it makes sense because the dataset for them was people reading books iirc

[marsh lagoon]

[glad nebula]

uploaded the contentvec version of legacy v2 #1408339243948052501 message

[mint elk]

does anyone recommend easeus voice changer? specifically for talking and realistic voices

[maiden tangle]

Hi guys am new to this AI world, but i want to become a master. Where do i start ? best free resources to learn ?

[sonic blade]

is there not a normal girl voice rather than anime bruh

[summer pecan]

There are free resources on YouTube. Check the School of Digital Marketing.

[summer pecan]

There are other AI tools that can help better

[hasty ruin]

Hello

[tiny cape]

You’re names look like mine but you are not AI

Look nice though

[radiant totem]

hi

[gleaming summit]

can someone give me sora 2 invite code

[sterile gale]

less attractive server tag, even though the previous server tag was good

I don't want to wear it anymore

[amber solar]

Hay

[blissful yoke]

Dose anybody got sore 2 codes

Sora*

[jagged verge]

I am looking for Linux based Voice Changer. I am using the soniTranlate. Is there any best ? If comes with docker image that would be perfect

[proud bone]

Hey guys - Also looking for a Sora2 code if anyone can help?

[lethal flax]

me 3 lol

also can you even get it on android or PC

[blissful yoke]

Please we are begging for our codes

[tame badger]

hi @jagged verge
there are a few good linux voice changers.
if you’re using soniTranslate, you can try Real Time Voice Cloning, so vits svc, or RVC they’re pretty solid.
some of them already have docker images too, makes setup easy.
you want something real time or for post-processing?

[covert lake]

I don't think anyone has sora 2 codes here

[blissful yoke]

If you dont got a code im gonna eat all of your tacos

[covert lake]

for help, please elaborate in #1192011222023950368

[jagged verge]

for now i want something based on file input. Can you share link?

[covert lake]

RVC, you can check it in the ai hub docs or use help channels

@chilly lake https://openai.com/index/openai-amd-strategic-partnership/

[chilly lake]

MX series, not consumer gpus lol

[covert lake]

I mean hopefully such unexpected partnership is going to benefit amd consumer gpus too

i expected them to partnership with nvidia instead

[tame badger]

sure
so-vits-svc-fork → https://github.com/voicepaw/so-vits-svc-fork
RVC (Retrieval-based Voice Conversion) → https://github.com/RVC-Project/Retrieval-based-Voice-Conversion
both support file input and have docker setups available.

[chilly lake]

not really, AMD has server cards and working linux drivers, this is basically trying to plug shortages of Nvidia cards for OpenAI

[polar flax]

-svc

[pine acorn]

https://media.discordapp.net/attachments/1159822429137412177/1194008385113301003/svc.gif?ex=67f2ddf6&is=67f18c76&hm=17bc260b884736770ef6374560c253ce24b348309d2ee35530100af590aef719&=&width=614&height=561

[tame badger]

wdym?

[chilly lake]

svc is ancient and outdated

[jagged verge]

Any other recommended by you? I want to host it on linux server

[wild tusk]

Any 1 got sora inv code plz

[chilly lake]

realtime?

[tired vine]

hiiiii

[hasty ruin]

Yooo

[fleet elbow]

I just have such a good idea

You all know minecraft

You know those ai generated 1980‘s dark fantasy minecraft videos on youtube or tiktok

Right?

So how about making an entire ai generated short movie (maybe not so long, but about 15-30 minutes, with ai generated dark-fantasy minecraft, wich give a kind of nostalgic vibe, maybe with pillagers who move with their troops to a dark forest with many creatures, then raid a village at night, with crossbows and axes and cannons and blow away the village walls with tnt, and set everything on fire and then steve comes and helps the villagers and its an epic battle in a retro dark fantasy style

[arctic zephyr]

ai carly

[wintry iron]

Who has built a cutting edge AI Agent and want to show off?

[uneven pine]

Sora 2 invite code????

[indigo wind]

What are formants

When I raise the amount it makes my voice higher pitch

[tidal sedge]

anybody have Sora 2 code? if you have i would love to get it 🙂

[indigo wind]

@gentle trench you’re right people do ask every day lol

[tidal sedge]

Fr

[gentle trench]

They do lol

[indigo wind]

did they just get banned x-x

[gentle trench]

No clue

I'm at work on lunch break

[indigo wind]

what are formants

[gentle trench]

Formants is like changing the pitch of the voice without making it sound like a smoker or breaking the way it sounds

Basically just a pitch shifter

[indigo wind]

Ahhh alr thank you

[chilly lake]

nope

not possible

RVC model actually removes the noise/background music/etc

[boreal vale]

openwebui + openrouter as provider is way too op ngl

[gentle trench]

don't waste your money on the impossible

[still sigil]

"use ai to remember your passwords" yeah thats a scam

definitely a scam

"i was looking for new ai tools"

nah man you made it

no self promo

you made it

you didnt find it

it doesnt "look" like a scam, it IS a scam

"use ai to remember your passwords" as if thats not a scam

either way, end of convo

no self promo

no self promo

even your little "everyone talking about it in reddit" knows its an ad

[quasi dome]

Chat who wants a sora 2 code

[trail island]

@quasi dome if you still have any I'm interested

[polar flax]

paid commissions/requests are not allowed

[polar flax]

@real pecan paid commissions/requests are not allowed. if you have issue on voice changer, pleast post in https://discord.com/channels/1159260121998827560/1192011222023950368 and explain it

[hardy fox]

i want

[hardy fox]

did u take it?

[slate sluice]

hey is there like a voice scrambler app?

or model

[real pecan]

Allrighty, thanks! Sorry about that. Its not really a problem with a voice changer, its more like a new feature, and The price point was mainly a motivator to see if anyone made this kind of thing before, or has an idea on how to do it. Thanks for removing it for me.

Probabbly missed the rule that stated that, leme refresh myself with the rules if I'm already at it

[polar flax]

the helper staff members here aren't selling any service to provide help/support, so it is free at all, not even accepting tippings

[glacial hamlet]

how do i get the rvc to sound more convincing? no matter what i do every model i have used so far sounds bad

[real pecan]

O-okay?

[polar flax]

you seem to be acting sus

[real pecan]

What?

Okay.

Have a good day.

[slate sluice]

how to fix the crackle in voices so it dosent sound like a smoker for 20 years

[lapis geode]

Could anyone send me a sora 2 code

[remote zealot]

where's my ai

[barren cliff]

Does anyone have a sora code lol

[restive hamlet]

just started using google colab to train rvc models

[covert lake]

please check #1402790586028789830 and use help channels

[covert lake]

if you need any help, use help channels

[gentle trench]

Go away shoo

[gentle trench]

[potent abyss]

hi guys, i wanna clone my voice to a voice model from this discord. i used to do it with huggingface spaces, but i cant find one anymore that doesnt give me an error when i try to upload a voice model link! is there any other ways where i can upload the voice models so i can clone them?

would apprecrtiate the help very very much ❤️

[steel frost]

does RVC have realtime voice changing?

[covert lake]

@steel frost @potent abyss
**Please check #1402790586028789830 and elaborate your help request in the dedicated channels :) **

[steel frost]

is a yes or no possible?

[glacial hamlet]

literally

[covert lake]

btw for any other helpers checking this convo, the user already got their answer #✨│ai-help message

[fervent sierra]

I need a Sora ai code pls

[gentle trench]

no

where do u people come from go away sora code people

shoo

https://cdn.discordapp.com/attachments/1401477083586101359/1419449901762478243/image0.gif

[molten crater]

hi

[last cradle]

Hello may someone help me with the femala AI voice changer?

ive seen a video from a guy called novision is this legit?

[pine bobcat]

hey guys, anyone used python t-string? I just checked it, it looks great. If someone played with it, can you share the exp.? I like to know from others.

[gentle trench]

The what, excuse me?

Get outta here if you're gonna call it that

No one will help you if you're egrill trolling it's against the rules here and actually breaking the law dummy

[last cradle]

what should i call it

[gentle trench]

That depends on what you're trying to use it for

Btw the help channel is here https://discord.com/channels/1159260121998827560/1159290139609137264

[last cradle]

to get rbx from e boys

good enough?

[gentle trench]

I'm so disappointed in you

https://tenor.com/view/oh-moderators-gif-25571015

[agile pewter]

someone can help me with sora code

[elder willow]

no

[gentle trench]

Oh yea <@&1159293140440723499>

Kill him

[elder willow]

Skin him alive and make his family watch

[gentle trench]

Yes

[worthy jewel]

wow

[worthy jewel]

pls adams

[unique pawn]

how do i reapply for model maker, i took a break from ai

ts changed alot

[restive hamlet]

YOU. I do not like Steal a Brainrot.

[covert lake]

We don't allow and won't help users for harmful ai usage such as financial fraud and catfishing (which is your case it seems, but also any type of misuse for bad/illegal things, since we don't want to be affiliated with anything bad), which is illegal, you have been warned

[restive hamlet]

can I PLEASE find a reasonable cloud alternative that isn't bombarded with ads and paywalls

[covert lake]

I don't think anyone here got a sora ai code NGL

Everyone's asking for it but it's rare and expensive to get

[covert lake]

#outdated-model-maker-role

[covert lake]

And yeah the server changed overtime

[unique pawn]

i dont have access

[restive hamlet]

y'know, for rvc voice conversion (huggingface spaces are one, i just need more)

[covert lake]

No cloud computing service will ever offer 24/7 free unlimited GPUs, they'd only lose money to buy all those GPUs and keep the server alive for free

[covert lake]

If you want to check for more cloud services, you can check the docs https://docs.aihub.gg/ and ask in the help channels, but you won't ever find a cloud GPU unlimited for free forever

[restive hamlet]

and i don't have a gpu on my laptop :(

[unique pawn]

how tf am i supposed to get model maker if its locked mannnn, someone end my suffering

[covert lake]

I mean I do understand that the main reason people use cloud is because they don't have / can't afford a GPU for their local PC, but at the same time no one will offer free unlimited PCs if u get what I meant

I wish I had an rtx 5090 and H200 too, but they are expensive afterall

[restive hamlet]

guess i'm sticking to huggingface spaces and google colab

[covert lake]

You didn't choose the RVC Enthusiastic role from the onboarding

We made 3 separate roles currently, one for the community, one for general ai users which you have, and one for rvc related channels

[covert lake]

There are also other cloud computing services with a free tier like Kaggle, Lightning.AI and Weights.com for RVC, you can check the docs

There are other free ways, just nothing unlimited, hopefully I explained myself clearly enough :)

[covert lake]

Also btw Kaggle applio or lightning.ai applio or weights.com would be more suggested for free tier

Google colab gives max free 4 hours Daily of GPU, which aren't granted so one day it can be 30 mins, the other 2 hours, the other 3 hours, etc

[covert lake]

Is openrouter that good?

[boreal vale]

Its 1 api key, you buy tokens, then you can use any AI model using a big list of providers including gemini, gpt 5 pro, deepseek, claude

I am unsure about pricing in general these days or if there are alternatives

[covert lake]

Ohh, I mostly use Gemini 2.5 pro or some local ollama model, WBU?

I mean do you actually use all that long list or mostly just some models?

[boreal vale]

I normally used chatgpt free before this but now with this setup I use deepseek v3.2 and claude because they are very cheap

gpt 5 pro is very expensive

[boreal vale]

no but its still nice to have

there are some ablated models aswell iirc

I just wanted a local ai frontend where I can build a knowledgebase so

[covert lake]

I heard many disliking gpt 5 tbh

[boreal vale]

deepseek and claude is pretty good for my use cases

let me check how much credits vs tokens

[covert lake]

What are your use cases if I may ask ?

I mostly just do researches and coding tbh

[boreal vale]

same, for coding i use it for configs, installation, small stuff, rough drafts, functions etc.

and I use it to learn coding via giving it a lua script I wrote and asking it to translate to rust or c++

it helps me understand it well

[boreal vale]

and yeah research ofc, It isnt my only source by any means but its good for pulling up quick information

or general direction

[covert lake]

yeah its always better to double check any info

there is misinfo everywhere nowdays

[boreal vale]

For example I ask questions like "What ball bearing is commonly available that is able to hold static load of around 100kg"

[covert lake]

just be sure it doesn't go crazy https://www.reddit.com/r/GeminiAI/comments/1lxqbxa/i_am_actually_terrified/

[covert lake]

smh reddit's embed sucks

but yeah ai can go crazy while coding

u go to the bathroom while it starts deleting ur whole database whoops

[boreal vale]

oh I never give the ai access to actual project

or, only read access

i do the writes

@covert lake with deepseek v3.2 exp, 97K tokens cost me $0.0285 USD worth of credits

thats not half bad tbh

[covert lake]

That's good for security

[covert lake]

Eh that's great

[boreal vale]

yeah comes to around 33 million tokens for 10 usd

30-35 around that ballpark

comes to approx 20-25

million words

[covert lake]

you used 33 million tokens?

[boreal vale]

no

i just calculated how much milage 10 usd would get me

[covert lake]

Ohh lol

most people dont usually need that many tokens

[boreal vale]

yeah and 10 usd is not that much

even 1 usd gets you a long way

[covert lake]

have u tried groq too btw?

not sure if it supports deepseek

[boreal vale]

i havent, let me try

[covert lake]

but it has a very generous free tier

[boreal vale]

wdym, I use openwebui as a frontend

[boreal vale]

i mean i can use openrouter to access its models

[covert lake]

https://console.groq.com/docs/models

it doesn't offer deepseek models

but it offers some few other models with good rate limits

[boreal vale]

i see but i already have all the grok models lol

openrouter is nice

[covert lake]

grok and groq are 2 different things, but yeah

[boreal vale]

ohhh ic

[covert lake]

i noticed it does sound similar 😭

[boreal vale]

anyway i gtg

[covert lake]

grok = elon musk ai model serie
groq cloud = a cloud API like open router with their own custom hardware, which is super fast and generous but doesnt support too many models

cya

[drifting sigil]

Hi.

I have a question

[main nest]

three hours later ... still no question.
Guess that was a lie then. ;-P

[distant ore]

Hey everyone, is anyone here working on AI automation projects ?

[elder willow]

ask your question

[drifting sigil]

How do I make planktons ai voice sound good with diamonds by Rihanna

[elder willow]

what software are you using?

[drifting sigil]

What software am I supposed to use. I'm not using a software Right now

[elder willow]

I mean like what ai are you using

or have you none yet

[drifting sigil]

None yet

But I'm just wondering

[elder willow]

ok i asked chat gpt and it said
"
Go to FakeYou.com, find Plankton, type the Diamonds lyrics, and download the voice. Then find a Diamonds karaoke or instrumental on YouTube and save it. Open Audacity (a free sound app), drag both files in, and move Plankton’s voice so it lines up with the music. Click his voice, go to Effects → Reverb to make it sound smoother, and Change Pitch if it sounds weird. Turn the music down a bit and the voice up until it sounds nice together. Then go to File → Export → MP3
"

I asked it to dumb it down a whole bunch if thst helps

[drifting sigil]

I still don't understand

[elder willow]

https://youtu.be/2udB-X1OUIE?si=veTrgm18euWwa441

hope that helps

if not wait for someone else who actually knows stuff abt ai to help you

[unique pawn]

since im back in the server, do i get my level back or is that something i gotta re-earn cause idk if i wanna re-earn like a year and a half ngl

[distant ore]

How can i use Sora 2 in germany

[unique pawn]

youll have to wait

sadly its not available for me yet either

[distant ore]

VPN ?

[unique pawn]

yeah that would work

but youd still need an invite code from the sora server

which is hard to get

[distant ore]

Ah okay thanks

[unique pawn]

ofcourse!

[distant ore]

Do you want to make money with it or are you just using it for fun?

[dense ginkgo]

Want to share what's your plan to make money ?

[elder willow]

im pretty sure you have to re-earn levels

you can always ask vijay

when hes online

[unique pawn]

gosh i hope not, thx ill ask

[elder willow]

good luck

[unique pawn]

tyy

[dawn temple]

We changed the levelling bot 😦

So all levels are reset

Since you are an old member, I have given you GIF and Media perms

[unique pawn]

tysm!

[civic echo]

how to make ai music ?

[zinc rune]

What hugging face sit do I use to convert a sound file into a different voice using the voice models

[drifting kettle]

Hii

[hearty sigil]

https://we-are-jammin.xyz/ please rate the animation

[north goblet]

guys how do i get sora 2 code

[gentle trench]

This is not the discord server for the code sorry

[maiden marsh]

can any1 elp me decide whether to go for chatgpt plus or gemini pro?

[paper tinsel]

anyone got the sora 2 invite code dm me pls

[shadow whale]

!colab

[pine acorn]

📒 Google Colab Notebooks

Google Colab is a Cloud (Remote Good PC) Service. While the Free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

• **Applio**

by IA Hispano
Google Colab

• **RVC Mainline**

by Hina
Google Colab

• **UVR5 NO UI**

by Eddy
Google Colab

• **UVR5 UI**

by Eddy
Google Colab

• **Wokada Deiteris Fork**

by Deiteris & Hina
Google Colab

• **Hina's Modified Original Wokada**

Google Colab

• **RVC-AI-Cover-Maker-WebUI**

by Shiro & Eddy
Google Colab

• **FaceFusion UI**

by Nick088
Google Colab

• **FaceFusion NO UI**

by Nick088
Google Colab

• **Music Source Separation Training (Inference)**

by Jarredou & Makidanye
Google Colab

[shadow whale]

no coalbs seem to work anymore

colabs*

[gentle trench]

Are you trying to train a model?

[turbid bane]

Have you tried the Kaggle one for voice changers?

[shadow whale]

Hey i'm just trying to make songs from ai voices

let me try it

where is kaggle

[potent cave]

does anyone have Sora 2 invite code?\

[frank dragon]

anyone do realistic models?

[neat saffron]

Anyone know how I can get a sora invite code?

[chilly lake]

Not here, pls dont beg

[neat saffron]

Nah not begging. Just simply asked a question, thanks anyways

[unique pawn]

How often do people review the model maker stuff now, cause I've been waiting for around a whole day just to get it 😭