flow-general・🌊

[copper harbor]

Remove the absurd invisible SynthID watermark.

[edgy holly]

Google Vids is a video editor. Had to be something other than regular Veo, with all the static text, promo layouts and such. Haven't used Pomelli yet, but the announcements on social look useful.

Suno is great. I started using it months ago thinking I could make some jazz/blues background music better than stock for one of my personal projects, and success on that front, but... it's so much better than that. Very fun.

[woeful tiger]

Oh snap, right. I know about Goggle Vids but haven't used it. My brain went straight to Veo lol

[vapid rampart]

Hi

[woeful tiger]

Hello..

[tawdry prawn]

this is funny, sam and dario not holding hands

[regal orchid]

this guy looks like harry potter

[bright schooner]

hey guys anyof you know why my flow? (google ultra) uploaded images for reference lately are not there? after i refresh the browser all recent uploaded images for reference are gone and only the old uploads are there

at first im like ok ill just reupload, but its kinda getting annoying now

anyone can help?

[dim robin]

Hi

[bitter aspen]

hi,
i have same issue.
i think first i will install money to that extension website.
second, i need to find API.
but where is that API ? how can i find.

[golden shadow]

I've had the same issue. I think it's a bug that's been around for a while.

[swift warren]

how can i use flow without credits

[ocean moon]

I’m having the same problem. Image uploads don’t work, and everything disappears when I press F5. It’s been like this for a week now, and the service is basically unusable. I’ve contacted support asking for a refund, but there’s been no response so far.

[humble ridge]

It'd be coool if flow will have the switch account button since some of us use personal accounts on work machine or vice versa. 🙂

[slate dock]

Bro doesn't speak like a gen alpha

[near palm]

why this happened a lots. its annoying

[turbid umbra]

not sure what's going on lately but VEO 3.1 Fast has been spitting out god awful generations, like its so annoying

they feel like veo 2

[bitter aspen]

yes?

[turbid basalt]

Hi everyone

I need help from someone who uses Flow

[wind saffron]

so now theyve changed the download process, no longer get a choice where we want the files to download to, well at least Im not anyway

[edgy holly]

I wonde if they might be running lower quality quantized versions during busy periods. The quality does seem to get significantly worse at times.

Prompt interpretation as well.

[bitter aspen]

if you ask me that did i fix my problem. no i didnt. i asked to the user "CAPCHA-VEO3" to be sure.

[river horizon]

how did u do that? please elaborate

[tepid wigeon]

Anyone else seen an issue with ingredients-to-video generations where it apparently generates an enhanced prompt behind the scenes and adds sound effects and other instructions, but then it generates a voice reading out those instructions or the name of the sound effect instead of actually following them?

[hollow pike]

Howdy

[weary marlin]

can any one confirm me? in veo ultra are there 25000 credits and in it veo 3.1 fast low priority uses 0 credits?

[ebon rover]

confirmed

[north wigeon]

Has anyone noticed a big decline in image quality and details when generating a image using gemini? I'm a ultra user but i prefer to use gemini instead of nano banana because i feel i can give it better instructions. Am i wrong about any of this?

[fossil marten]

I noticed the same thing on Gemini Nano and the worst is on veo3.1 the logic for example for stages of construction of a construction site is clearly degraded since last week yet I use the same detailed prompt structures

[north wigeon]

These dips in quality are very annoying. I feel the quality never fully recovered from the last dip.... and now this on top.

[timber river]

Carefully analyze the full video and break it into logical cinematic scenes.
A Important Rules: Each scene must be exactly 8 seconds. If a sequence is longer, split it into multiple 8-second segments. Maintain story continuity between scenes.
Do not skip emotional transitions.
A Formatting Rule (Very Important) Each generated video prompt must start with: Prompt 1: Prompt 2: Prompt 3: And so on. Each prompt line starting with "Prompt N:" will be parsed. Do not add extra text before the prompt line. For Each 8-Second Scene Provide:
1️⃣ Full Cinematic Breakdown Include: What is happening in the story during these 8 seconds Character details (age, clothing, hairstyle, expressions, body posture) Emotional state of each character Environment and location details Background elements (weather, objects, vehicles. effects, etc.) Camera angle (wide, close-up, aerial, tracking, etc.) Camera movement (push-in, pan, tilt, handheld, slow motion, etc.) Lighting style (dramatic, soft, neon, stormy, golden hour, etc.) Color tone and visual mood.

2️⃣ Generate: One Highly Detailed 8-Second Cinematic Video Prompt The prompt must: Be written as a continuous paragraph Describe active motion (not a static image) Include camera movement Include lighting dynamics Include emotional intensity Be visually immersive Be production-ready Represent exactly 8 seconds Be 4K cinematic quality Style Guidelines High-detail animated cinematic style Dramatic immersive lighting Strong emotional storytelling Ultra-detailed environments Smooth character motion Dynamic camera work 4K ultra HD

[fading dirge]

veo just failing everything today?

(Close-up of a man at an outdoor backyard prep station generously sprinkling coarse kosher salt all over a massive tomahawk steak, no talking.) like whats the reason this prompt keeps failing? like bruh

[unreal belfry]

seems like it

[fading dirge]

yeah i think so too

[unreal belfry]

so sick of flow not working. i pay good money to use this and i make money with it. when it doesn't work, i can't create, if i can't create i can't post and if i can't post i don't get paid. anyone here know of ai video generating that is better than this?

[fading dirge]

well honestly right now their are a couple of generators that are better, but veo has unlimited and speed, thats what makes it competetive

[unreal belfry]

yeah, everything i've tried is slooooow

[fading dirge]

grok is pretty fast but at 480p, i havent tried it at high res, but grok is right now better than veo 3

[crystal talon]

Veo is also ultra woke

[fossil marten]

What is strange is that the quality is degraded in a few days, the model has become very lazy, I find

[unreal belfry]

how so M.E.G.?

[crystal talon]

Well, when i type a man walks down the street in newyork, they are mostly indian or asian. Worst thing was, i prompted 3 soldiers run away from an explosion in the jungle, it gave me 3 female black soldiers. I mean, it is more than obvious. The fact that we cant animate any woman in bikini is also a joke. Literally yesterday, i typed father and daughter arguing. Bro, it gave me a black father and a white blonde asian daughter. Like what? 🙂 @unreal belfry

Btw in here i cant even type w t f , this tells a lot.

[spark jungle]

Feedback: The "Treasure Hunt" for Value (New Subscriber Journey)
I recently subscribed to Google AI Pro for video generation, but the user experience is broken and confusing.

1. The "One-Stop" Fallacy:
   I subscribed to Gemini Advanced expecting a creative suite. Instead, I hit a "3 videos/day" limit immediately. It was never communicated that I needed to leave the main app and go to "Google Labs/Flow" to access my real allowance (1,000 credits).
2. The Hidden Wallet:
   I had to ask an AI multiple times to understand that "ImageFX" uses a daily quota while "Flow" uses monthly paid credits. This "split wallet" isn't visible in my account settings. I almost cancelled because I didn't know the credits existed.
3. Regional & Top-Up Dead Ends:
   As a Canadian, Flow gives me a 404 Error. I'm paying for credits I can't use. I also calculated that buying credits à la carte is cheaper ($25 for 2.5k), but discovered I can't access the store without an active subscription.
   Fix: Please unify the dashboard. Show my 1,000 Labs Credits inside the Gemini App settings. Don't make me hunt for a separate website just to use the features I'm paying for.

Feedback: Why Google is Losing Creator Mindshare (vs. Midjourney/OpenArt)
I read reviews and watch tutorials constantly (Midjourney, OpenArt, Leonardo, CapCut). Until today, I had never heard of "Flow" or "Veo."

1. The Visibility Problem:
   The only model name I recognized was "Nano Banana." In the creator economy, your naming convention is invisible. I actually confused your video tool with "Flux" (Black Forest Labs). If a paid subscriber can't tell the difference, the branding is failing.
2. The Dashboard Gap:
   Competitors like OpenArt have visual dashboards (galleries, folders, history). Google offers either a "Chat Box" (Gemini) or a "Developer Console" (AI Studio). Neither feels like a creative studio. "Flow" is the closest, but it is so hard to find it might as well not exist.
3. Value Prop:
   You are hiding your best value proposition (high-volume generation in Flow) behind a wall of silence. I almost cancelled because the "3 videos/day" limit in the main app felt like a bad deal.
   Fix: Stop treating "Flow" as a secret experiment. If it's the core value of the "AI Premium" subscription, it needs to be front-and-center with a unified "Creator Studio" dashboard.

[slow yarrow]

yo how do i get my hands on veo 3 cause i’ve been trying and i cant

[full flicker]

Hi please share your feedback in the #1238621511401734254 channel. The team monitors those threads and it will also ensure your feedback does not get lost in the chat here.

[edgy holly]

Gemini, the Flow web app (https://flow.google) and all the AI video platforms that have multiple models.

[slow yarrow]

k

[umbral fjord]

agree with the degrade in quality! tiny details look like painted using a brush whenever I generate on Veo

[crisp basalt]

Gemini uses nano banana pro model tho, idk what you mean by you use Gemini model instead of nano banana-

[north wigeon]

I mean i ask gemini to create the image instead of doing it through flow

[crisp basalt]

In Gemini, it’s like the user prompt gets a “auto expansion” before it creates an image or video. This happens behind the scenes, which is why it usually gives great results (though it can vary depending on how complex the prompt is). Flow, on the other hand, lets you tweak your prompt more, so if you don’t write it just right, it might not do what you want

[north wigeon]

can i use reference images in flow? ive never been able to get it to do what i want when it comes to images. When generating with gemini, i give it 3 to 4 reference images of characters in with the same background and setting just the character is slightly altered...gemini then creates me a new variation of the character when i ask for it. It flow i cant seem to get that to work

[crisp basalt]

There’s an official documentation for this. Let me try to find it.

[north wigeon]

o wow that would be a huge help.. thank you ❤️

[crisp basalt]

https://ai.google.dev/gemini-api/docs/image-generation#template here you go

sorry for late response ,

Each section has templates that you can copy and use.

[crisp wasp]

everytime i upload iamge in nano banana pro its show voilate our policies can someone help?

[plucky elbow]

HI im new here

need to know why this keep repeating if i put an image , and if i jv script direct the image can genereate

[crisp wasp]

happens with me same

[summer vapor]

Hi everyone — if I generated a video in Flow with a voice I really like, what’s the best way to keep that exact same voice across all future clips?
I’m making multi-shot videos, and I need consistent voice tone, pacing, and character identity from clip to clip.
Do you recommend generating silent clips first and then using external TTS + lip sync, or is there a reliable in-Flow method for voice consistency?

[clever axle]

I have to say that Flow improved a lot, i stop getting random prompt blocks

[serene cradle]

Why can't I upload any image to flow ?
Anyone got the same ?

[timber fable]

/imagine

[near gate]

It would be cool if we 'flow users' had a community where we supported each other, discord doesn't seem like the place though its way too broad and people usually only come here to ask a question and bounce.

[crisp basalt]

I used to help people here whenever I felt like it, but lately, I’ve been avoiding trying to be helpful. Most people here are rude for no apparent reason.

We used to have community challenge for flow too

[near gate]

I wasn't trying to be rude, I have been hanging out here for a few months and just watching as people come and go and realizing there is such a huge amount of us making videos with VEO that we aren't competition we can actually enhance each others work.

[crisp basalt]

I wasn’t..even saying you’re being rude

😬

[near gate]

Just being clear because I often times have problems getting my message across with text

[crisp basalt]

I’ve been here since video-fx era

You’re fine lol. I was just surprised a bit

[near gate]

Snow got my brain going slow today, the blizzard rolled in.

[crisp basalt]

Usually, #flow-tips-n-tricks・✏️ is meant for sharing tips and tricks, but I’m not sure if it’s because of a language barrier or if people simply don’t want to understand or read the rules before posting anything, so they just post wherever they want

Rest in peace #1319423780057251921 community challenge channel

[near gate]

Google just has way too many different little programs that do too similar of stuff so it divides the communities too, like whisk and nano banana / veo are very very similar programs. Whisk has some stuff that I wish Flow had and vice versa

[crisp basalt]

Oh trust me they’re gonna sunset whisk and image fx soon

I can’t say much more, but yall should definitely check out this year’s Google I/O. 🙂 It’s going to be packed with new and interesting stuff, especially for Flow

I hope they would launch a community-based website similar to Flow TV, where users can showcase their video generation. The labs team would organize occasional challenges, and even feature users’ creations on the front page

Now why am I typing like ChatGPT lmao , ew

[steady pebble]

Hi,why my account is not eligible for using flow? I'm in a eligible country

[simple shoal]

Has Nano Pro's quality been having issues for anyone else here? I'm not sure how to get it to make 2k - 4k images anymore

[crystal talon]

Why doesnt flow veo3.1 have a CPP? They dont need partners I guess?

[turbid umbra]

high demand ? cant generate

[unreal viper]

Hello, as a friendly reminder we cannot generate images or videos in the server! Please do not post your prompts in the channels as it will not generate. If asking for advice or tips on your prompts please use #flow-tips-n-tricks・✏️ ! Thank you!

[misty burrow]

is the veo3-generate API returning a Code 14 error for anyone else?

[umbral jasper]

for me it says high demand, i cant generate any videos with google veo3.1 rn

[misty burrow]

thanks for confirming brotha -- thats Code 14 btw

[twin mortar]

Glad I'm paying for Ultra 🙄

[glossy island]

Me too

[crisp basalt]

i think google generative ai server is having outage or something

[honest wren]

I just paid for Plus just to generate a 5sec video and got the same issue.

[crisp basalt]

u can generate 5 sec vid on Flow?

i thought it can only do 8

[edgy holly]

I'm the 50th follower 🙂

[main saddle]

Flow's on Instagram! check us out 👍

[crisp basalt]

not anymore!

nvm

[ebon rover]

who else is getting Something went wrong... constantly? clear cookies it works for 1-2 prompts then again same error

[exotic crystal]

Yes, I have this problem too!

[unreal viper]

Hi, videos/images aren't generated in the discord server. In order to make a video you'll need to go to https://labs.google/ pick the tool you want to work with then go from there. If you need help with prompts please ask in the appropriate channels for advice. Thank you!

[silver escarp]

Hi i cant get to my projects anymore. Anyone else having this issue?

[broken hearth]

What is the capybara from the Instagram video referring to when it mentions an 'exciting update' it'll be sharing with us soon?👀

[ivory gazelle]

Hi, Can I use the image I already have on my PC as a character? if so please tell me how

[analog lava]

@exotic crystal Please create a thread in our #1238621511401734254 detailing your issue. We ask for your patience as the Flow Team work through the numerous reports. Thank You

[honest wren]

I don't know I prompted to do 5sec but it did nothing, 8sec is ok for me as well.

[vast wolf]

Hello guys

[crystal talon]

I watched clips of mine from oktober 2025, veo3 was much better back then. What happened?

[strong mountain]

I just want to say that Flow is absolutely amazing.

[cosmic charm]

I've seen many videos of this skeleton character in hypothetical scenarios generated by Google Veo and it's very cool. Gonna make one of my own

[broken notch]

May I know when is the time to generate images again after my 100?

[proud wyvern]

@tacit halo
hi

[elfin brook]

hi, am beginner. how to start FLOW

[exotic crystal]

@analog lava - I understand what you're trying to do here, but when someone is paying $250/mo for a service crucial to business devlierables that isn't working, time is money. For every day I'm being asked to wait, is Google going to refund me for that loss of time? If so, let me know how to request a refund.

[compact phoenix]

@tacit halo
Hi Google Labs team,

I am consistently getting a "Failed Generation" error on the Flow UI. Upon checking the developer console, the API (aisandbox-pa.googleapis.com) is returning a 403 Forbidden error, which appears to be a reCAPTCHA block.

I am using a legitimate Google Workspace account on a clean network. Could you please investigate and fix this backend issue? It is completely blocking my workflow.

[quartz kraken]

its not working here either. are the servers down?

[compact beacon]

hey guys

cant use flow...

getting an error that flow cant generate voice at the moment. any solutions?

any solution guys?

Someone from the support, please give it a look. its my 2nd day using flow and gets my work behind. No time for bugs.... Please have a look at it and solve it. Thank you.

[compact beacon]

any news on flow?

@compact phoenix @quartz kraken you managed to generate any video on flow? i am still getting errors.

[compact phoenix]

@compact beacon @quartz kraken If you clear your browser cookies and log in again, it will work temporarily. But after generating a few times, the error just comes back. It's only a temporary workaround, we still need a real fix from Google.

[proper star]

/imagine prompt: 3D rendered minimalist Kedarnath temple flood, floating ground cut section, glowing cracks, isometric view, unreal engine, 4k

/imagine prompt: 3D rendered minimalist Kedarnath temple flood, floating ground cut section, glowing cracks, isometric view, unreal engine, 4k

[still prism]

Anyone getting robbed on credits? Mine has been deducting double even on failed generations. I'm yet to see my failed generations refunded also.

FLOW taking double credits

[unreal viper]

please can all be reminded of our first rule: Be compassionate and treat others with kindness and respect. we have #1238621511401734254 for feedback specific questions/ queries

[ebon rover]

day 3 of flow being pretty much unusable for me.. Something went wrong. error constantly. clear cookies and it works for 1-2 prompts, then same again. repeat. anyone else?? not cool on ultra plan to deal with this

[fluid blaze]

I have the same thing

appeared a few days ago

[ebon rover]

any clue what the problem might be? i have friends who do not have this problem and it does not seem widespread

[ebon rover]

i dmed you lets figure it out

[umbral jasper]

are you still having issues as well ?

[solemn solstice]

after paying high price still having issues with video generation specially in scene builder 2 hours work taking 6 hours because of errors worlds most popular company still cant make tools stable what a shame

also its using user gpu for upscale and still have daily upscale limits

[lyric osprey]

Wish google would communicate with us better. so many days without being able to generate a veo 3.1, audio sync issue or somthing after %50.. Ive got work to do

[onyx cedar]

Agreed — clearer status updates would help a lot. Even just a timeline or acknowledgment of the issue would make planning easier.

[umbral jasper]

I thought it was back earlier today bc it let me make some generations

but seems like not

[spiral python]

When update ?

[crisp basalt]

👀

🌊 🌊 🌊 🌊

[night verge]

Hi everyone! Please help. When I go into Flow and click the “Create with Flow” button, the page reloads and doesn’t let me access my account (workspace). I’m using a corporate VPN, and everything was working fine until today. What could be the problem?

[quasi harbor]

Hi everyone, do you have any ideas how to handle sfx + VO smoothly for videos? I tried elevenlabs but it felt like a punch to my ears, completely lacking emotion

[devout lagoon]

can i get some free credits on flow?

[midnight kestrel]

hello, where can I ask about nano bannana issues?

[stark cosmos]

Hi everyone! Does anyone know what Is the limit reset time for nanobanana pro in Brazil? Almost 12 hours limited here. Also this Is really a terrible costumer service i'll say. Is It so difficult to add "you reached your daily usage, come back in xx:xx"? So i have to wonder and try every hour to see if i can work on my project again?

[lyric dust]

If you have the budget, you can try Google AI Studio's TTS first to design voice styles, or the Qwen3 TTS solution. I use the latter because it's cheaper

Just the voiceover part

[wind spoke]

Every single generation I make fails this morning. Come on dude. Its like you tightened the restrictions like crazy overnight. I'm trying to make a living here and the second I find a format that works- you guys make everything impermissable.

[placid dragon]

guys its my first time using flow
i want to animate a logo
is there guid to make prompt
i have the logo as image

[quasi harbor]

I spent an amount on Google TTS, and it seems to support English well, but it didn't work as well for my native language, Vietnamese, it feels a bit creepy.

[quasi harbor]

If you prioritize quality, using only one prompt is not enough.

Try sending logo image to Gemini and ask it to suggest image-text-to-video prompt versions based on that

[lyric dust]

I spent an amount on Google TTS, and it

[lyric dust]

https://cloud.google.com/blog/products/ai-machine-learning/ultimate-prompting-guide-for-veo-3-1
https://ai.google.dev/gemini-api/docs/video?hl=zh-cn&example=dialogue#prompt-guide
As KOBK mentioned, try using Gemini; however, you need to understand some of Veo's basic logic, otherwise you'll be misled by the AI

[quartz kraken]

have there been any ackknowledge from Google about the issues?

[arctic kelp]

Hi team,

Recently, many generated videos are coming out completely black. This issue has been happening for a long time and still hasn’t been fixed. Is anyone else experiencing the same problem?

I hope the Flow team can investigate and resolve this as soon as possible. Currently, credits are only refunded when there is no audio or when the video fails to generate. However, when the output is a black screen video, credits are not refunded, which results in a significant waste of credits.

Please look into this. Thank you.

[spare rapids]

Why does Flow suddenly look so different?

[opaque sand]

Flow's design has just been updated.
I won't say that Flow's design and interface logic have changed and that's why I don't like it.
I will say that it is no longer intuitive to use and the interface is all crooked and skewed....
Or does anyone like it?

[crisp basalt]

AHHH

ITS FINALLY HERE

[fierce nimbus]

Lots of improvements all over the place!

[crisp basalt]

@fierce nimbus @viscid mesa thank you for this wonderful update ❤️

tho i been meaning to ask this,, what happened to jumpto and prompt expansion feature?

[wide ether]

Whoa! Big update!!! I can see you can download videos in batches, but that still doesn't solve the problem if you have hundreds upon hundreds of videos to download ☹️

[fierce nimbus]

how can we improve

[crisp basalt]

i think that can be doable by addin all those generated vid into new collection

[viscid mesa]

these are not present for now, but let us know if you miss them! we'd love to hear

[wide ether]

do like you can do in whisk: download all videos in one operation .... better yet, have option to download all favorited videos

[viscid mesa]

you can use the left rail to get to all your favorited content and multi select to download them all together 🙂

[wide ether]

There's no disable on the auto playing of videos when you hover over them

[versed peak]

what in the world is going on with the update??? I am getting all these unnecessary random objects being added to my scenes now, especially in the background, how do I keep it consistent now

[fierce nimbus]

no model changes with update

[fierce nimbus]

you can set precference for auto sound on hover or silent playback on hover.

[crisp basalt]

@wide ether hi i just find out about this just now lmao

[versed peak]

But Im using the exact same prompt ive used over the past 30 minutes.. and since the update hit, huge change in consistency. No changes to prompt, or reference image.

[wide ether]

yeah I saw that for the sound ... but it still wants to play the video

[fierce nimbus]

yup. Videos play on hover to distinguish from Image and seek faster

[wide ether]

OMG!!!! that helps! thanks!!!

[wide ether]

ah!

[strange forum]

Did the layout change? Or did I do something.. I like the old one

[versed peak]

I have a scene with sunglasses, before the update, they never changed, not in 50 generations.. but since the update, the shape of the sunglasses is always different in each scene....

[fierce nimbus]

are you still using the same Veo model? Model picker may have changed Veo model.

[wide ether]

FWIW - I do like the update .... and it appears that the loading of videos as you scroll down through your list of videos is really optimized now!! Nice job on whoever sped up the querying and presenting as you scroll through the list of your videos!!!

[wide ether]

Software engineer here ... so I really notice these things 🙂

[fierce nimbus]

performance was a MASSIVE focus of this update.

[wide ether]

... and it really shows! kudos!!!

[fierce nimbus]

collections (folders), drag and drop, @ tagging, new asset drawer, marquee selection, 2 view modes (grid/batch), WAYYYY snappier to load, navigate, etc... and WAY more to come

[crisp basalt]

just a suggestion, i would love if theres some kind of button that can do "auto describe/auto name" assets /ingredients

[viscid mesa]

anything new that you create will be automatically named! but this is awesome to hear for your old assets

[modest lotus]

@fierce nimbuswhat is wrong ? i cant generate any video .. all stock at 99

new update

[fierce nimbus]

lemme check capacity with team

[versed peak]

Veo 3.1 - fast.. the only difference is that it's now auto-selecting ingredients and that seems to be changing consistency. Cant get a scene with the same sunglasses now ugh

[fierce nimbus]

What do you mean by auto selecting? Like draggin in to prompt box? @ tag?

[versed peak]

Before the update, ingredients was a separate button, now you can't opt out of ingredients

[modest lotus]

@fierce nimbus hey .. did they remove add to frame ?

[crisp basalt]

are u trying to do text to video? if so , just dont add any images/ingredients to your prompt box and it should do text to video

[fierce nimbus]

ah you just dont have to use them. NOTHING changed model wise. We collapsed ingredients to video and text to video in the UI but nothing backend changed

[fierce nimbus]

frames can be found in setting panel. Please give us feedback here.

[versed peak]

But I'm telling you, I've been doing the same prompt (image to video) for the last 30 minutes before the update. The sunglasses and background were conssitent. Then the update hit and all my scenes have some small addition to the background (like a lamp post or an extra car) and the hair is slightly different, or the sunglasses change. It's losing the consistency for some reason.

[modest lotus]

i mean if i want to take a frame from the video inside the flow ?

[crisp basalt]

oh that one

[fierce nimbus]

right clikc > save as frame

[sage hull]

there was a model nno banana flash what was that, now its gone

[toxic shale]

How can i get out of the large grid view?

[toxic shale]

I have to click on each vid to hear the audio

[fierce nimbus]

settings top right. You can choose size and grid/batch

[toxic shale]

Thank you!

[fierce nimbus]

This update aimed to make Flow endlessly more flexible while retaining the features you love (like batch view)

[versed peak]

Oh I figured it out!! Phew ok

[fierce nimbus]

My favorite update is Collections. I can organize assets. I can make a new one and just prompt in there and then back out and have a clean top folder. I can make one for each character im building. Collections can house collections

[hazy obsidian]

the flow aint producing videos now

[crisp basalt]

ah i guess generated asset/ingredient doesnt do auto naming?

[modest lotus]

i can't even save a frame

[fierce nimbus]

For new assets we should have auto names. You can also name any asset (uplaoded or generated) by tapping top left and writing your own name

[fierce nimbus]

let me report. Is the save frame function showing? just not saving?

[modest lotus]

yeah but im making a kids video.. could be that the problem

it was working normaly before this update

[hazy obsidian]

I'm facing this for a while now after the update @fierce nimbus

[fierce nimbus]

for a while meaning the last hour or for a while meanign multiple days? Whats happening here

[hazy obsidian]

for like 20+ mins

[modest lotus]

you guys also deleted so much frames from my account

[storm vault]

Can flow creat or generate without background (png) ?

[warm orbit]

Quality now costs 200 credits?🫥 was the actual model updated tho? Or it’s just the interface?

[gritty stream]

its so easy to tell the new ui is made with gemini, but it geniuely looks insanely good so W

[turbid umbra]

I cannot see the history of pictures I generated anymore? 🙁

[gritty stream]

welp they didnt run a codex review on it 🤣

[viscid mesa]

this drawer is now specific to a project! if you want to see your old ingredients, we added a project called "Ingredient Archive" where you can access these and copy them into other projects

[modest lotus]

same

[viscid mesa]

credit costs are the same! we updated the interface to show you how much you're spending at once rather than just per generation

[turbid umbra]

where can I find this?

[viscid mesa]

it should be in your projects on the home page!

[turbid umbra]

checking

[turbid umbra]

apologies, cant find it, could you send a SS?

[tacit plaza]

You can create a png animation with a green screen background and remove it in any editor (like use ultra key effect in premiere pro)

[hazy obsidian]

is this supposed to happen since the Interface just got updated? 😭

[dry oar]

Minors bugs I’m sure will get fixed but I’m liking the overall look

[analog radish]

Guys, there are bugs in the new UI. For example, when we reuse a prompt, the reference images should be populated. This is a pretty big issue.

Also, the gallery is pretty hard to work with.

[edgy holly]

You can create individual images using Nano Banana in Flow, ask for a blank background, then delete them in an image editor. No transparency.

[warm orbit]

Ah my bad, now i see!

[latent jetty]

WHERE ARE ALL MY FRAMES

[surreal hare]

How long veo3.2 or veo4 will available?

[unreal viper]

Your creations are safe! In the coming weeks, you'll see an opt-in button appear in Whisk [/ ImageFX] that lets you migrate your library directly to Flow. You can also download your content at any time before April 30, 2026. After that date, any remaining content will be permanently deleted.

[latent jetty]

bro i dont use whisk where are my frames from FLOW

the new ai got rid of them

[turbid umbra]

Ok will take a while to get used to the new UI, loving it so far, I like that it's snappier.Have you considered a Desktop app? maybe to unlock more capabilities.

[analog radish]

Also, the videos need to be stacked somehow. When we’re testing different prompts, we need to know which stack of four the prompt belongs to.

[latent jetty]

this is ridicuous the prices i pay to have this happen to me

[analog radish]

I like the fast deletion feature. It would be great if we could have it for non-failed videos as well, but with a confirmation dialog, please.

[turbid umbra]

the play on hover is chef's kiss love it

[sick sparrow]

pohhh finally the drag and drop for new images! Thanks ❤️

[languid olive]

how do i make it go back to normal i cant see my prompts in a giant list anymore wahhh

[latent jetty]

the new ui is trash

[languid olive]

not there

[viscid mesa]

sorry about this! it should be there within the hour!

[languid olive]

this makes me wanna rip my hair out

[viscid mesa]

try out batch mode in the settings panel on the top right!

[analog radish]

On the wheel in the upper right you will find "Batch or Grid" the default is Grid

[latent jetty]

WHERE ARE MY FRAMES

[turbid umbra]

its ok, just downloaded from other and uploaded, experimenting now with the Camera Motion

[latent jetty]

WHY HAVE THEY DELETED ALL MY ASSETS FROM THE OLD UI

[turbid umbra]

they are going to be back soon

[latent jetty]

BUT WHY HAVE THEY DISAPPEARED NOW

[analog radish]

Theywill appear

[latent jetty]

THE PRICES WE PAY FOR THIS IS RIDICULOUS

[remote bay]

Hello, as stated in ⁠#start-here・📍, “Currently, our team can support engagement in English; the discussion in this server will largely remain in English for now. Thank you!

[languid olive]

what is this this is so bad ui

[pallid violet]

Hey, all your creations are safe! The UI is new and some bugs can be expected, but the team is working hard to fix everything asap. Don't worry, nothing is deleted!

[turbid umbra]

gonna be cool soon guys, they gotta sort out the kinks steady

[vocal bridge]

hey everyone, how do i get unlimited nano banana pro for free for lifetime locally?

[languid olive]

just give us the default view back please

[analog radish]

It’s not that bad. Give it a try and create a couple of generations.

[viscid mesa]

everything is still rolling out! apologies for the kinks but we should be there within the hour 🙂

[languid olive]

all these new models come out and google wastes its time breaking the ui , instead of dropping a veo upgrade =[

[analog radish]

Despite a couple of bugs, the new UI has some awesome features.

[latent jetty]

REAL

[turbid umbra]

would've been nice if you left the option to toggle to legacy for a while, but I guess since its labs thing didnt worth the effort i get it.

[burnt oriole]

we have 4k upscale now?

[onyx eagle]

The new ui is totally wrong. SO HORRIBLE!

[summer rapids]

What did you just do ? We can't extract pictures from generated video anymore in scene builder. You just ruin all my workflow.

[burnt oriole]

[vocal bridge]

so what is the meaning of this forward? why did u send it now?

[fallen citrus]

Nice work with UI, guys, i see a lot of improvements. I'd be nice to be able to copy the prompt with the images instantly. In the editor i am not seeing that feature.

[primal cradle]

I cannot access the new flow. It throws me a 404.

[latent jetty]

new ui is bad #bringbackoldui

[remote bay]

We hear you, and we're sorry this news is frustrating. We know Whisk/ ImageFX means a lot to you. We made this decision to give your creativity a better, more powerful long-term home in Flow — and we're committed to making the transition as smooth as possible. We hope you'll give Flow a try!

[latent jetty]

HASTAG IF IT AINT BROKE DONT FIX IT

[main saddle]

this is a forwarded message from a user within the community. I forwarded it as it was in a channel that is not monitored by the Flow team. thanks!

[languid olive]

i almost rarely whine about this type of stuff im a power user in higgsfield , i teach a course, i teach google flow !! you guys just ruined my entire workflow , i absolutely despise this ui , oh and now theres a default setting that deletes our prompts seriouslt who is working in this department that thought oh hey lemme just screw up the entire ui with 0 announcement or warning this is ridiculous

[burnt oriole]

I tried the new experience with Flow and here are my feedback:

• firstly, I noticed that when I tried to paste an image on the prompt box (by CTRL V), then it gets uploaded but also appears as a generated image along with them. I’m not sure if it’s a bug or if it’s supposed to happen like that.
• One of the things I’ve noticed when generating multiple videos simultaneously for a multiple scene video is that I loose track of the sequence of the video. I wish there was a way to number them so I can arrange them properly in the scenebuilder.
• I noticed that the User Interface isn’t optimized for android.
• Bug: when I generate multiple clip after clicking on a clip: then if i try to add that to project, it doesn’t appear in the canvas, instead only the first clip appears in the canvas.

Some suggestions on the UX:

• I wish there was an option for collections to be added as ingredients. Especially when we have to handle multiple ingredients which are same across various different videos/images, it becomes hard to always keep adding them so an option for collections could be helpful as it could add them all in one click given that they are distinctly stored already.
• I noticed there is a very beautiful video preview when I click on a video, i wish there was an option to add that video to scenebuilder from that preview directly. It could be a minor improvement but could make the process fastened.
• adding skip and previous button in the video preview would be great to check multiple generated videos one by one.
• I wish I could see variations of image within the media canvas. I noticed that the images are stacked and can be only viewed and used after clicking on the image.
• It would be better if the option for “Add to prompt” could be on the menu that appears when I hover on the media. (Would also solve the problem I mentioned in first prompt). Also it would be better if the “multiple videos” if generated in bulk could be stacked instead of edited images

[viscid mesa]

you can do this in any video editor! simply open up the video, play it, and right click on the frame you want to save

[burnt oriole]

I also shared it via the in built feedback feature

[void ledge]

It is surprising that there is no option to create square images

[languid olive]

everything is out of order ....

[languid olive]

can we please do a roll back

[unreal viper]

We hear you, and we're sorry this news is frustrating. We know Whisk [/ ImageFX] means a lot to you. We made this decision to give your creativity a better, more powerful long-term home in Flow — and we're committed to making the transition as smooth as possible. We hope you'll give Flow a try!

[burnt oriole]

i think it would also be great for the 1:1 size images as whisk and imageFX are going away

[viscid mesa]

we're looking into it! in the meantime, you can crop your images to square in the editors

[summer rapids]

I do baby's adventure. I can't upload children picture but I could export frame from generated video. It's been 3 months since you said you should extend this to nanobana and you just do the opposite there

[latent jetty]

right whos deleting my messages

[remote bay]

Thanks for the feedback! Please send any feedback or bugs reports to #1238621511401734254. This is so we have it all organized and it won't get lost 😄

[latent jetty]

limiting my freedom of speech ridiculous clearly they dont like feedback

[burnt oriole]

ooww sureee. sorry i posted it here :))

[glossy sparrow]

whisk features like style arent available for now in flow?

[primal cradle]

I cannot even access the new flow. So I cannot say nothing 🙁

[remote bay]

absolutely no worries 😄

[main saddle]

we're investigating this!

[languid olive]

i dont care about whisk i care about flow i care about this i care that i house all my projects in this ui i care that i save all my prompts here like a library i care that yall just nuked it

[summer rapids]

I just can't work anymore cause you broked the only working option needed

[primal cradle]

Ok

[latent jetty]

new ui is bad

[languid olive]

at least you cant see the new ui ruining everything

[turbid umbra]

gonna be back soon they said in an archive feature

[latent jetty]

watch them delete my message

[languid olive]

an archive feature ??

[turbid umbra]

yea

[languid olive]

what even is that

i just want a normal view feature

[glossy sparrow]

can someone send new ui

[void ledge]

This is a bad option. For some reason, AI always does a poor job of drawing full-length human figures in horizontal images.

[glossy sparrow]

dont have access to it rn

[summer rapids]

New UI should add new things, not delete old working workflow

[turbid umbra]

Camera Motion working for anyone? mine all failed

[main saddle]

we're looking into why some members don't have access to the new UI!

[primal cradle]

I don't know, man. When I could see what's going then I'll realize XD Then I tell you my opinion

[main saddle]

hey y'all: access to the new Flow UI is still rolling out, so hang tight if you're not seeing it yet.

you should also be seeing an Ingredients Archive folder with your assets / frames showing up in a bit.

[little fulcrum]

Hi my generations are all stuck on 99%. When I refresh they are gone. 🙁

[fallen citrus]

@main saddle congrats on new UI. Is there a way for previous prompts to be added instantly as in previous Flow ? It really speds up the process of generating when multiple generations of the same prompt are needed till the model gets it right.

[summer rapids]

How can we access the old UI so we can continue working and not being restricted by option you deleted?

[primal cradle]

So I'd have to wait for a moment? I ask because I'm not in a hurry 🙂

[viscid mesa]

you can adjust your settings to not clear the prompt on submit! each tile should have a re-use prompt button in the hover menu

[main saddle]

yes, that's what I'm aware of! feel free to ping me if you're not seeing the new UI by EOD.

[tacit plaza]

It is there,you can resuse prompt

[languid olive]

scroll up i posted ss

[primal cradle]

Ok. Take your time! Thanks 🙂

[fallen citrus]

When creating images? I dont seee the option. The history of edits appear on the right, but the only buttons are download, save to project and flag

[fierce nimbus]

switch it to batch mode and its 95% old UI

[main saddle]

fwd from @turbid umbra

[rugged glacier]

10 Credits relly ?

[rugged glacier]

in my country or is it still displayed for me

[turbid umbra]

thanks, my bad thought I was here 🙂

[main saddle]

all g!!

[little fulcrum]

I can't get any generations done. They are getting stuck at 99%

[burnt oriole]

hi guys what should I make with flow today 🙂 I wanna try something new!

[rugged glacier]

[edgy holly]

Smart

[main saddle]

hi Annie! thanks for reporting this, will take a look - please feel free to open a ticket in #1238621511401734254 as well

[edgy holly]

Rollouts don't happen all at once. I didn't have it earlier. Now I do.

[turbid umbra]

yea yea, got the workaround for now export reimport from other projects.

[rugged glacier]

I wish Lipsnk Studio had voice control and motion capture, and face capture 3.2.

[edgy holly]

I like this a LOT

[tacit aspen]

Question -- is Expand gone? I'm not seeing it.

[tacit plaza]

Click on info symbol and you will get the option to reuse prompt

[visual juniper]

Click video

[edgy holly]

Right click

[visual juniper]

YOU CAN DO THAT?

[edgy holly]

Nano Banana is MUCH better than Whisk at style transfer.

[turbid umbra]

no deletion of extensions available?

[turbid umbra]

you have to start from scratch now if you dont like the 3rd or 4th extension.

[summer rapids]

There is no option to extract frame in scene builder. You can't right click.

There was a +

[turbid umbra]

yup that too

[primal cradle]

@main saddle I could access the new flow. Thank you! 🙂

[main saddle]

yay thanks for your patience! 🙏

[summer rapids]

Now, we can't reuse pictures with children from flow videos even the boss from flow said that

[viscid mesa]

this is now available on any video! simply click on it to open up the editor and right click on the frame you want to save

[summer rapids]

So give us back that option at least

[primal cradle]

You're welcome!

[summer rapids]

It block due to minor in video

[winter brook]

Hi Annie, which model are you using?

[novel plank]

Are images and videos from projects getting ported to Flow for us from Whisk or do we need to do it individually?

[viscid mesa]

There's a banner available on Whisk that allows you to do this! It won't happen automatically

[edgy holly]

In scenebuilder, I can.

[vernal remnant]

how do I download in batch? I do not see that option to select videos or pictures to download by selecting them in 'batch' option?

[viscid mesa]

click and drag (or use shift click) to select multiple items, then right click any item to download them all in a ZIP file

[summer rapids]

There are no children in your video. Was able to extract frame before. Now I have this

[proper drum]

Suggestion for the UI (its generated with Nano Banana Pro in this image, but looks better than the actual)

[wicked crater]

I can't even generate any image with a referance.

[vernal remnant]

I see, it says, "we cannot handle dropping videos at this time"

[turbid umbra]

anyone else can see the seconds counter?

on timeline

[viscid mesa]

you should still have the download option on right click!

[main saddle]

for those of you who are looking for your previously generated Ingredients (Images), we added a project called "Ingredient Archive" where you can still access them.

please let us know if any of you are still having issues!

[edgy holly]

Their filters get better with each update.

[vernal remnant]

yes, but that is one at a time, not in batch downalod to download multiple videos at once

[summer rapids]

Here is what the flow boss said

They just done the opposite

[turbid umbra]

did anyone find the archive yet?

[viscid mesa]

there isn't an archive in this update but let us know if you'd want to see this

[kind bison]

All of these UI changes, yet the core video model has fallen behind all of the competition and the censorship is worse than in Communist China.

[edgy holly]

I don't know what you mean by archive, but the Library has never worked well, going back over a year. Use something off the shelf if you have to. Or let us save to our Gdrive and search there.

[rustic lagoon]

i write normal prompt and it stills says it violates community guidelines, hahahahaha this is wild

[edgy holly]

I don't work for Google. I'm just telling you the pattern. Every major update, the filters get a bit better at catching content, and kids are one of the things they filter.

[proper drum]

What model of Image 4 we have available?

imagen-4.0-generate-001
imagen-4.0-ultra-generate-001
imagen-4.0-fast-generate-001

and why is only 1k output?

Also Nano Banana Pro is 1k generated and available 2k and 4k upscale? no 2k and 4k generated?

BTW: I'm an Ultra subscriber, not Pro or Plus (without discriminating against users on lower-tier plans).

[summer rapids]

They filter kids not generated inside flow. Why I can expand a video with a kid and not extract a frame to create a new video from it. It's complete nonsense.

[proper drum]

Another suggestions.. could the UI have a way to add more selections? like designs softwares

[manic sierra]

Feedback: Regressive UI Change & iPad Usability Concerns

I am posting this here because the recent UI update has made Flow Lab almost unusable for my workflow. I want to emphasize that I work exclusively on an iPad (no PC or laptop available), and the new interface is a massive step backward for tablet users.

Main Issues:

• iPad Navigation & Saving Broken: The new UI has completely broken the native "Long-press to save" functionality. Previously, I could long-press an image to bring up the iOS system menu and "Save to Photos". Now, long-pressing does nothing. The "Download" button is a very cumbersome workaround on iPadOS.
• Loss of Iterative Workflow: I used to rely on having the prompt visible under the image with a one-click button to reload/edit it for a new generation. Now, the "Chat/Edit" logic forces me to overwrite my previous work instead of creating independent versions for comparison.
• Hidden Information: Having to dig through "i" icons or sub-menus just to find and copy my own prompt is extremely tedious on a tablet.
• "Edit" vs. "New": The tool now feels like a restrictive chatbot ("What do you want to change?") rather than a professional creative suite that allows for precise prompt engineering.

Conclusion:
For an iPad-only user, this update offers zero advantages and only adds frustration. It feels like the creative process is being throttled.

Request:
Please restore the "Classic Interface" option or bring back the one-click "Use as Prompt" functionality and native iOS image interaction. A professional tool must remain accessible and efficient for tablet-based creators!

[wicked crater]

I see why, the ingredients is only on video generation.

[rustic lagoon]

my opinion is that this update redesign was completely unnecessary, very bad.

[onyx eagle]

Please do the roll back. This isn't right. so uncomfortable.

[summer rapids]

And we can't expand video with 4 generation at a time.... It's limited to 1 at a time and stacked on the old video. So much counter productive

[remote bay]

Thanks for the feedback! Please send any feedback or bugs reports to ⁠#1238621511401734254. This is so we have it all organized and it won't get lost 😄

[viscid mesa]

you can still generate 4 at a time! simple click on the selector in the prompt box to adjust the number of generations per prompt

[summer rapids]

no option when expanding,it's only one

Just searching workarounds until you re-enable the option to extract frame with kids, but even that is more limited than before

[turbid umbra]

the problem is that it does not carry over in the extend feature, once you are beyond the first generation.

[viscid mesa]

ah yup! for anything in the editors it is one at a time now - let us know if you think it's still useful to have multiple versions here!

[turbid umbra]

its massively useful! I wrote in the feedback, it cuts generation time a lot!

[vernal remnant]

I'm still trying to find the option to batch download images with upscale, I dont think there is a way.

[turbid umbra]

I might not like the first extend, so I 100% of the time generate 4 extends and pick the best.

[vernal kernel]

Hey, why did they botch the flow ui? Cant use it on mobile no more, jesus

[quiet wadi]

what´s the matter with flow
Flow has changed it´s interface, I can´t find my image resources now

[winged plinth]

Why this Hugo angry😂

[manic sierra]

Question regarding the new UI / iPad Workflow:

Could a moderator or admin please explain how to achieve a "One-Tap Prompt Carryover" in the new interface on iPad?

In the previous version, I could simply tap a button to load the prompt of an already generated image back into the input field to refine it and generate a new, separate image.

Now, when I tap an image, it opens in "Edit mode" and seems to only allow modifying the existing image. How can I quickly reuse a prompt from my history to create a fresh generation without manually copying/pasting or overwriting my previous work?

Since I am an iPad-only user, I am looking for a touch-efficient way to do this. Thank you!

[quiet wadi]

It´s Flow got a new actualization=

?, I can´t find my Image resources

[harsh kiln]

isso ficou uma porcariaaaaaa

[winged plinth]

On the top left corner.

[quiet wadi]

But there appears the images I´ve created not the images resources I have uploaded

[viscid mesa]

these are available in a new project on your home page called "Ingredient Archive" - right click any of them to copy and paste them into the project where you want to use it!

[royal laurel]

plz keep whisk,, flow isn't work the way whisk work. whisk has so many features flow hasn't

[quiet wadi]

But until now I could use the image resources I had been uploading , now all the resources go out

[harsh kiln]

@viscid mesa now VO3 are not following the prompt structions....

[wicked crater]

Oh and there's no option for square size.

[royal laurel]

which site is this?

[wicked crater]

Flow

[sinful lotus]

We used to be able to extend a scene and then go back to the main list and export 1080p for each of the 8 seconds.... now it combines them and only 720p is possible to be selected. Am I missing something. Quite a setback if not able to download the 1080p for each of the extended scenes...hopefully I'm being stupid?

[quiet wadi]

I not understand what the *** is going on with the new Flow update

all my resources are not on my collection

[turbid umbra]

got the same, cant find the upscale anymore

[shut olive]

Is there a way to revert to the previous version?😭

[quiet wadi]

your resource´s library go out too?

[fierce nimbus]

three dot menu. Download. Choose option

[shut olive]

@quiet wadi if you are talking about the uploaded photos. yes there is a mess. everything is gone. only old uploads

[quiet wadi]

So, now must upload it again all of them

[sinful lotus]

Same problem... all prior resources lost...

(the assets/photos I uploaded)

[sinful lotus]

It is there for 8 second clips, but if you extend, the original clips got put into one bigger clip and you no longer have a 1080p download option.

(that I can see anywhere)

[quiet wadi]

Probably they did it intencionally

[edgy holly]

"Loss of Iterative Workflow: I used to rely on having the prompt visible under the image with a one-click button to reload/edit it for a new generation. Now, the "Chat/Edit" logic forces me to overwrite my previous work instead of creating independent versions for comparison."

This is a big deal. Haven't had time to play with it yet, but if I have to go through some convoluted mess of clicks to iterate, that's a HUGE waste of time.

[viscid mesa]

wanted to call that you can still access / use your old versions with the history panel! simply click on your favorite to save it as the version that is used, or hover over in the panel and save it as its own tile

[forest plover]

Wondeing if anyone else is seeing this issue and if theres anyone I can reach out too on the flow support team for help. I went to generate a video with 3.1 and it generated 20-25 outputs at once - when I specified 2-4. I was charged the credits as my balance went down (not sure what it was when I was generating) but I refilled and lost anothe 8000 credits after that. Attached some screenshots here. (I also got banned for some reason last time I posted). If someone on the admin team can reach out, please help!!

[edgy holly]

Nano Banana does everything Whisk does, but is much better at it.

[edgy holly]

What is a history panel?

[quiet wadi]

Then we have to upload the images everytime we create a new project

[harsh kiln]

The same prompt and the same initial frame image are generating completely different results in VO3.
It seems that VO3 is responding differently to prompts. Has there been any change to that?

[quiet wadi]

I really don´t know to use the new Flow

my images don´t be saved on my library

[viscid mesa]

when you open up any image or video, there's a panel showcasing the editing chanel on the right side

[viscid mesa]

we haven't changed anything on the model side with this update

[summer rapids]

Can you give an option to have the old UI so we still can work with our working workflow. It's not fair to delete or remove options like that

New UI should be better, not the opposite. Maybe there are improvements, but all I can see is that I can't work anymore

[fierce nimbus]

whats not there for you?

[edgy holly]

I gotta actually do stuff on the clock today, gaaahh

Okay, so I can't speak for the other guy, but that's not what I'm talking about. I'm talking about being able to access past prompts without clicking around.

If you're picky and/or have high standards, to get what you want you almost always need to pull up one of the last prompts several times so you can run it again, for me usually with some alterations. That's easy to do with the 'Batch' view mode, and how it was yesterday, since those prompts are right there on the right (which is good for me since I like that format), but in Grid you'd have to go through a few clicks to access older prompts. A prompt history should be right there around the input box.

[summer rapids]

Can't extract frame with children from Flow generated video (in order to continue video from that frame or generate other version to get a good one from that good frame) like it was before the update and like your boss said it should work even on other tools. Not only it's not working on other tools but doesn't work anymore in flow.

[quiet wadi]

It´s possible to return to old Flow?

[summer rapids]

And also the 4 generation when expanding and the fact that they are now stack inside the initial video. Some said we can't also download 1080p versions of those extands. Didn't tested that yet

[royal laurel]

in flow Image ratio is a big problem.

there is only two options, portrait and landscape.. but whisk has more

[strange forum]

Is there away to go back to the old interface… this ain’t it

[clever axle]

awesome changes on flow

can u roll back?

scene build is dead, and the extend is not working 🙁

[quiet wadi]

the images are really bad

with that new update

[modest tinsel]

i dont know who to work whit the image in the flow

[open goblet]

is there a way to revert flow to the old interface?

[sonic zealot]

I think its best to have your frames and generated images in a separate folder outside of flow

[clever axle]

the way it display is not bad, but is not very intuitive, some videos generated do not even appear

[sonic zealot]

I've had that issue recently. Usually when I refresh it appears

[fierce nimbus]

whats not working. You can choose settings top right and choose batch mode to bring interface 95% of the ay to old Flow but there is not a way to revert

[brisk dew]

I can't even delete the image.

[winged plinth]

You can right-clicking your favorite or preferred clip and selecting Reuse Prompt, it will instantly populate in the prompt box.

[clever axle]

extend is not working how it was suppose to work

[open goblet]

is it possible to make the batch downloader use 2k as default?

[clever axle]

you cant also save the last frame as a new image

[strange forum]

Seem like a lot of people don’t like it.. idk just ain’t it man lol and also on phone is not good also it doesn’t copy the image… idk man just not feeling it

[brisk dew]

God, thank you that I didn't start a new project.

[clever axle]

cant remove the videos I disliked from the collection

[winged plinth]

Right-Mouse/delete

[quartz juniper]

Why can't I drag a file to another tab now? This new "version" is awful!

[clever axle]

not this type of collection , but when a vid has 3 layers icon on upper right sidr

[brisk dew]

I used a tablet for work.

[open goblet]

With all respect to the devs - then i feel these UI changes was made to slow down our workflow deliberately, so we would use less resources.

[pliant shuttle]

Hi I'm just a beginning ai content creator. I just found out about flow. I can already tell I'm going to lov3 this way of creating

[quiet wadi]

I can´t use my image resources that I had uploading along this months

[clever axle]

same

[brisk dew]

I was looking at Freepik today. I liked their node structure. The developers did everything to make me pay attention to other platforms.😩

[manic sierra]

I totally agree with you. For me as an iPad user, it’s even more frustrating. What used to be a seamless process is now a mess of menus. It really disrupts the creative flow when you have high standards and need to iterate quickly.

At least you got a reply from a moderator, even if it didn't solve the problem – my questions about the iPad workflow were just ignored completely. I really hope they realize that this 'Grid' logic is a step backward for the overall user experience.

[summer rapids]

As an exemple I can still "add to prompt" pictures from minors when they are generated from nanobanana (like in the old UI) and generate video from it. It's frame with minors from Flow video that I can't now add to prompt' (it was working on the old UI so it should still work and not give the "not allowed to upload minors" cause these are from generated video.

Seems like you're using an download/upload rather that using the frame directly so it's blocked. I think you can revert back to the old option to use frame preventing the blocking

[quiet wadi]

but it´s possible to back to the old version

?

[summer rapids]

no

I now have that with frame from Flow video

[strange forum]

Well I hope they listen to the community maybe have an option for new and old.. it’s worth a shot I guess

[quiet wadi]

but I don´t understand anything of that new version, I can´t locate my image resources library , and the images now are so bad

[winged plinth]

You mean inside the iteration collection [Extend, Remove, Camera]? You're probably right, removing unwanted iterations will clean up the workflow a bit. [+1] vote for that.
That said, the new UI feels much more accessible to me, and I think with a few tweaks based on community feedback, it will meet expectations.

[manic sierra]

Many people have asked this already, and I’m also burning to know the answer. It’s a crucial question for my workflow: Is there any way to roll back to the previous version? A clear answer on this would be greatly appreciated.

[clever axle]

you also should take a look at the extend, it is working in a weird way.

i dont have much credits, so i am going to wait for tomorrow

[unborn ocean]

https://www.youtube.com/watch?v=uPVlPxcLFv8&t=87s

[summer rapids]

I confirm you can't Export to 1080p or 4k extended clips like we can before

[quiet wadi]

but confirm if is possible to return back to the old version

[viscid mesa]

try dragging them to the top left!

[viscid mesa]

we're looking into this and working on it!

[clever axle]

let me try

nope.. sorry Anika

[dry zenith]

The size limit for this new prompt box is absolute garbage... the old UI expanded higher with more text, now I can barely see 8 lines of text in the prompt...

[vernal kernel]

Why do I need to click again to download the video after upscaling, come on guys

[dry zenith]

👆

[strange forum]

Anyone using an iPhone? Can you copy the image?

[vernal kernel]

I am sorry for being negative, but this does not currently feel like an upgrade to my personal user experience.

[quiet wadi]

Google must pay us an issurance for that shiit

[hexed moss]

Do you also sometimes get videos that generate completely black with no sound?

[open goblet]

yes happens to me very often

[hexed moss]

annoying

[open goblet]

i mentioned it to support but they just said they were aware of it

[hexed moss]

alr

[quiet wadi]

you have noticed that the quality of the images are worst¿?

[open goblet]

did you change it back to nano banana pro? after the new update it was reverted to the normal mode for me

[quiet wadi]

How I could do that

please tell me

[sinful lotus]

I have been having a lot of trouble today with VEO 3.1 staying locked on reference photos. Sometimes great, sometimes completely ignoring.... maybe a coincidence?

[main saddle]

can I have a screenshot? 🙏

[vernal kernel]

Oh, that was the issue, I was wondering too, that the quality had dropped

[quiet wadi]

I have alredy put nanobananapro

[open goblet]

its easy to overlook as well.

[vernal kernel]

Yeah, I was on my phone as well before, so was very confusing. I did find it now, as I came to my PC

[sand pond]

• The website is slow, please remove the background demo videos.
• Also there are some dummy projects that have no thumbnails, content or title, this is a bug.

[quiet wadi]

It´s no way to back to the old flow¿?

[open goblet]

I have had the same issue about 5-10 times, so here is a screenshot 😄

[toxic shale]

Is it just me or are the images from NBP that get downloaded dont have a file icon image its just the PNG logo instead of a little preview of what the image is, its making organising the images quite tedious

[strange forum]

It’s not just you

[toxic shale]

im having to open each file again and again to see what is what :')

[quiet wadi]

but is possible or not to return the old flow?

[thick shadow]

Tried the new layout and when making a test video, the generations failed. Not a fan so far.

[hexed moss]

I deleted the one that was bugged :/

[hexed moss]

but kinda same as here

[toxic shale]

I dont think that would be possible its not an offline app thing where you could just go back a version by downloading a version this is the standard version now

[wicked crater]

Meaning it has to be Nano Banana in order to use ingredients.

[low axle]

Sad but true. I don't like this new flow

[quiet wadi]

until now my generated image promp convert automatically to english

now rest on spanish

[lavish field]

the new flow is not bad guys once you navigate around it

[open goblet]

it has a few neat features but for image generation its a step backwards

[quiet wadi]

any more have that issue ¿?

[low axle]

Can anyone tell me how to upload the image frame to make a video? I don't see an option

[gray lynx]

Hi folks. Two questions please. 1. How do I save a generated video frame now to "assets" and 2. how do I add to frames in scene builder from my assets now that the "add to frame" plus sign option is gone?

[summer rapids]

Right click on a video. But won't work if there is a child in the frame anymore

[gray lynx]

right click on the video in scene builder?

[summer rapids]

In the video when you view it. Not in scene builder. They remove that useful option

[winged plinth]

Click on "start" then "upload"

[low axle]

I can't see this option

[gray lynx]

Thank you for the information. It is soo sad now to see the extra all over the place items compared to the UI before when it was so much simpler to ue then

[restive stag]

Hi there, sorry for disturbing. How can i use flow here on discord? Thanks

[wicked crater]

Oh and 50 credits daily might generate 2 videos per day as well...

[desert berry]

I have not tried the new update yet, it looks like many are not happy, lets see

[open goblet]

do you get 50 free credits daily?

[wicked crater]

Yes. I'm still on the FREE plan.

[open goblet]

lol i got the pro plan and you get more credits a month than me then 😂

[wicked crater]

For some reason:

[dusty aspen]

Hi everyone, I noticed that after migrating to Flow, I lost the option to see an automatic description of the imported image. This is very helpful in understanding how the image was created or how I can generate variations of it. Does anyone know where I can find this in Flow?

[fierce nimbus]

There is no ability to rollback. What is not working.

[gray lynx]

Sorry another question. When I right click the generated video I can get it to go to scene builder which is fine, however how do I also add that video to assets to reference it to the next frame? I tried downlloading the video but when I try an upload the video from my hardrive to assets, the file is greyed out and I cannot upload that reference video.

[sand pond]

Video and image thumbnails need to be visually distinct.

• The thumbnail previews could be labeled with unique icons, such as Nano Banana, Veo, Imagen. (Veo and Imagen need logos first!)

[limber blaze]

Zoom out to 90%. Mine didn't show by default either.

[summer rapids]

So, if no roll back possible wiill you re-add the ability to extract frame from flow video with minor to use it as start frame or do I need to search alternatives right now to do my job ?

[winged plinth]

If you mean extending it with another video generation, simply open the clip, single or double-click on it, and from there you'll find the Extend option, which uses the original video as a reference and continues seamlessly from that point.

[manic sierra]

As I detailed in my previous posts above (and as the other user also perfectly described regarding the 'Loss of Iterative Workflow'), the main issue is the removal of the one-click prompt reload and the cumbersome menu navigation on iPad.

It’s not that a specific button is 'broken', it’s that the new UI logic itself makes the creative process much slower and less intuitive compared to the previous version. Please take a look at my earlier comments for the full context.

[viscid mesa]

our team is working on a fix for this now!

[summer nebula]

Please I can't upload photos here

[winged plinth]

You can do so by clicking the Upload button here.

[summer nebula]

I have uploaded it there, but the results are different from the photo I provided.

[solar pecan]

help-me

I have the premium but I can't use the veo 3

[gray lynx]

TY

[quiet wadi]

But you cannot save your uploaded images if you open a new project right?

[toxic shale]

When generating videos its altering some things from the start frame e.g. the name of the brand on a product changes or the logo changes. Does anyone know whatsup?

[dry zenith]

new VS old prompting boxes

didn't see a channel for feedback - guess it was removed lmfao

For having to continually go back into prompts and micro-adjust hundreds of times this is unacceptable to not have an option for the prompt boxes to expand vertically to display more of your prompt

[main saddle]

#1238621511401734254. you have to pick which channels you'd like to have in your channel list, it's in the "Labs Support" category! thanks for providing your feedback 👍

[viscid mesa]

this is really useful feedback! our team is looking into how we can make this better!

[strange forum]

I can’t download image it’s greyed out

[wicked crater]

Oh and Imagen 4 is not doing ingredients (subjects). Only nano banana.

[strange forum]

Using a iPhone 16 pro max

[onyx eagle]

Worst update ever

[grand trout]

😮

[latent jetty]

Wish I could pin this comment

[latent jetty]

Correct which is ridiculous, I’ve lost all my frames and I can’t find them anywhere on the website. Shame on Google

[fierce nimbus]

you an also just drag and drop anyhting from desktop into the tool

[fierce nimbus]

What can we do better. Whats working/not working for you @latent jetty

[fierce nimbus]

working on it. wE found the issue causing this. Sorry about that one

[solar pecan]

I have the premium but I can't use the veo 3

[remote iron]

How do I get a refund for my first month's subscription? Why wasn't I told that all videos created include a watermark? Please help me. Thank you.

[latent jetty]

Main problem and I’m sure people will back me up on this is “if it isn’t broken, don’t fix it” why change the entire ui just for things to go wrong and break like having all my asset frames that I use for frames to video be deleted and other issues people are having. It’s really hard to understand why. And it’s incredibly frustrating to learn to navigate a whole new ui when I use flow on a regular. The whole thing is just ridiculous.

[quiet wadi]

is there a platform where we can explain any google service or department the issues we are having with that update

[low axle]

Same issue

[low axle]

💯

I still can't figure out how to upload image.

[viscid mesa]

try the + button in the top right!

[low axle]

There is no button

[umbral fjord]

Loving the new UI. I can work much faster now.

[viscid mesa]

top right of the whole page! near your profile icon

[dusty aspen]

Hi everyone, I noticed that after migrating to Flow, I lost the option to see an automatic description of the imported image. This is very helpful in understanding how the image was created or how I can generate variations of it. Does anyone know where I can find this in Flow?

[prisma ferry]

i want to recreate/edit a still i created in flow but two of the reference images are greyed out. I assume they are ones I created in flow and aren't linking due to the update. this update has destroyed months of work, I trusted all my work would be safe. there is nothing better about this. pointless, the ui switching between images and video worked really nicely and was all clearly laid out, noone asked for this

[north wigeon]

Im finding hard to keep track of videos with the new smaller layout

[quiet wadi]

It´s not true, I lost all my resource images on my library , you can´t remik the images you make

I really need to back to the old Flow

[viscid mesa]

we noticed this too! to help out we added the "batch view" in settings and the option to filter by videos on the left side -- let us know if these help!

[viscid mesa]

greyed out images means that the model you have selected is not compatible (there should be an explanation when you hover over them) - does switching to Veo3.1 fast fix the problem?

[north wigeon]

Ah yes, this batch view helps a ton. Thanks ❤️

[viscid mesa]

do you have the "ingredient archive" project? they should all still be there for you to use and copy into new projects 🙂

[quiet wadi]

how could I found it?

[viscid mesa]

if you go back to the home page with all of your different projects, this project should now be there!

[north wigeon]

One more thing that would help is if you can watch the video with audio without having to go into full screen like before. im noticing i have to go fullscreen for every video when i want to view it. Is there also a fix for this @viscid mesa ?

[viscid mesa]

yup there is! if you toggle "sound on hover" in the settings in the top right, you can have the audio play directly from the grid when you hover over a video

[north wigeon]

thanks! is there a way to also make the viedos a bit larger in batch view?

[quiet wadi]

I cannot see it

[viscid mesa]

yes! the Small Medium Large grid settings in that same panel should help out here!

[north wigeon]

Thanks anika!

[prisma ferry]

im not sure what you mean? the only model available is nanobanapro. i used 3 reference images to create this image i want to recreate with edit to the text but 2 of the images are no longer there, i'm not exactly sure what they were and will have to test. Is my library of uploaded images still available its not there when i press the plus button. this is crazy to mess with peoples work

[viscid mesa]

this is not working as intended, we are so sorry about that! investigating now!

[prisma ferry]

thank you

[quiet wadi]

the ingredient archive must be there on the home page but where

[devout lake]

wow im so loving the new update !

finally a prompt that is not 3 line tall lmfao

[prisma ferry]

about 10% of my created images don't have the 'reuse' prompt button available

[viscid mesa]

do any of these images have edit stacks by any chance? because the edited image didn't have the original prompt we disabled it in that case

[devout lake]

collections are so good ! seems to be more responsive and not laggy as before when going through creations

[quiet wadi]

ok I locate it but how I can transfer the images of the Ingredient Archive to the new project

[devout lake]

did you put more RAM google in ur servers lol ? generation time with 3,1 fast low prio seems fast than before

[viscid mesa]

right click on any item to copy it, and then enter a new project and right click on the empty space to paste it in!

[quiet wadi]

but in the old version you have saved all our resources when you had started a nw project, you didn´t need to do that

[prisma ferry]

yes most likely I usually generate 4 options at a time. why disable this? it's needed daily, I always chose the prompt to reuse based on the quality of the image not having to trawl through everything to the source. Where are all my uploaded images please? they seem to be missing.

[viscid mesa]

in this version, we wanted to make each project cleaner and easier to work in - we're hearing that this is a pain point and are looking into ways to make it better!

[viscid mesa]

your uploaded images should be in the "ingredient archive" project!

[quiet wadi]

No you knew that was more usefull have all the images when you start a new project

[devout lake]

can u download multiple creation in 1080p at same time ?

[prisma ferry]

ok thanks found that project but they aren't useful to me in a seperate project. When I hit the plus button (for a video or image) that used to take me to my uploaded ingredients they aren't there and seem unaccessible. so are they no longer useable as reference images for generations in other projects? i choose most used and nothing

[viscid mesa]

we totally hear that this is a pain point and want to make it better! in the meantime, you can copy them over to use them in any project

[unique abyss]

The new U/X is super fast and responsive. I’ve been traveling, and was able to get some useful things done, even on my smartphone (even under the bandwidth constraints of using roaming data on high speed rail at 300 KPH.).

Before, I wouldn’t use flow on mobile other than to occasionally download shots for YouTube shorts, now I can quickly add some special effects to a photo I snapped, and share it immediately with my companions.

I was also able to start and finish a longer project on my Chromebook, something I probably would only have attempted on my desktop rig before the new updates, and improvement to ingredient organization and prompting.

I agree that it will be nice if the team could make reusing prompts more efficient, but regardless, this is a significant upgrade for the way I work.

[prisma ferry]

what do you mean by copy over to use in any project? how do you copy over within the platform? do you mean copy individual images into the project so uploaded images are mixed in with generated images and then drag into a prompt rather than use the plus button to access uploaded images or do you just mean download and reupload when I need to use in another project? I cant believe the team have done this, is it a mistake thats being worked on or was this the desired workflow?

[viscid mesa]

right click on any item to copy it, and then enter a new project and right click on the empty space to paste it in!

[tacit aspen]

I'm still having an issue with expand - where a prompt would get a little re-write to give it some additional help. Is that feature gone?

[prisma ferry]

this is really bad workflow. is it a mistake that will be fixed so I can easily access my uploaded images with the plus button as before or is this the intended new workflow?

[viscid mesa]

this is the intended workflow but we hear that it's difficult so we are looking into how we can improve it as soon as possible!

[prisma ferry]

improving it would be just having it exactly as it was. I can't believe this made it to the consumer, its made it unuseable. Please consider going back to the previous UI, thanks for your assistance

[cold inlet]

@all

[hazy kestrel]

jesus maria macarena. is there any way to get the old flow back? i can't work with this new update. horrible

[frigid void]

Did they disable flow upscaling?

[forest ridge]

With the new UI, it seems like when you extend a video, you can’t upscale the extended version afterward. Is that correct, or am I just not seeing the option?

[hazy kestrel]

before i could just hover over the pic, the drag button would appear and i can drag back and forth. now i have to click the stupid video, wait a second for it to load in a new window just to scroll back and forth. i can't understand the thought process of ppl who approve updates like this. like seriously?!

i create dozens of videos in minutes, i dont have time for this

my complete workflow is gone. before i was fast, now i am clicking around like a beginner who uses the tool the first time

auto-download also doesn't work anymore smfh

i could cry

[winged plinth]

Right-click context menu now supports key commands including copy, send to prompt, reuse prompt, ...etc. It's a notable UX improvement over the previous version.

[simple stone]

goodbye Whisk 👋
will new Flow has a feature to create a prompt from uploaded image like in Whisk? I like this feature so much!

[frozen tiger]

Hi everyone! I hadn’t used Flow for two weeks, and when I logged back in I was surprised by the new changes to the interface (among other things). At first glance it looked interesting, but then I ran into the unpleasant surprise that the new videos I’ve generated as extensions of others can’t be scaled to 1080! Is this really how it’s going to stay? Is it a bug, or am I missing something? It really ruins the video I’ve been working on for months. 😢

[viscid mesa]

we're sorry for the inconvenience! we're working to bring this back as quickly as we can

[frozen tiger]

I really appreciate it!, hope to get it soon 🙂

[thick shadow]

So, was this tested at all? Or are we the testers?

[little fulcrum]

Is there any change FLOW will be up and running again today? I haven't been able to work all day

[fervent condor]

Did we lose 4k images? I can only download 1k

[marsh inlet]

Does image generation in Flow not support a 1:1 aspect ratio? I rely heavily on this ratio for my work. Since Whisk is being migrated to Flow, I'm concerned that the 1:1 ratio support might be lost in the transition.

[opal citrus]

As alternative, there is option to crop it to square

ISSUE: UI

Search icon cut off on mobile

[arctic kelp]

😔

[marsh inlet]

Yes, I am aware of that, but there is a significant difference between a native 1:1 ratio and a landscape one. When an image is generated natively in 1:1, the AI adjusts the entire composition and framing to fit that specific square ratio.

[surreal hare]

Hi why button expand prompt gone for new function?

[viscid mesa]

4k should still be there when you click download on an image! note that this is only possible for images generated with Nano Banana Pro

[frigid void]

I have the Pro version, but even so, only the 1k option is available.

@viscid mesa can you help me?

[viscid mesa]

was this image generated with nano banana pro? can you share which model was used?

[frigid void]

Oh, sorry, I was silly. Flow was defaulting to the regular Nano Banana card and not the Pro version. I changed it and it worked. Sorry.

[viscid mesa]

End of day check-in: Thank you so much for trying out the new Flow UI and sharing your feedback candidly as we polish the edges!

The team is still working on several papercut fixes and looking into past ingredients not showing up for some folks - we'll keep updating you as we progress.

Keep your feedback and suggestions coming in #1238621511401734254 and #flow-general・🌊

[umbral fjord]

oh sorry that happened. I did not experience that on my end. Everything is intact. And the new UI has definitely helped with ny workflow. This is just my experience so far. Everyone else of course would have a different experience.

[umbral fjord]

Please add the ability to batch download assets too - if it already is implemented can you guide me through it please. Thanks

[viscid mesa]

you should be able to click on the empty space and drag to multi select assests, then you can right click on any selected asset to download them all

[chilly epoch]

What happened to flow guys ? I can't create even a single video.

[charred tendon]

Hello! I really enjoy Flow and have been using it for several months. These new changes look like they have affected the general workflow. There no longer is the capability to take screenshots in scene builder. I also don't see extend or any options to build off clips in scene builder.

imo taking screenshots was a powerful tool so a little surprised that feature has been removed

atm I don't see a way of extending clips with reference files

[viscid mesa]

hey there! both of these features are still available but on the single clip level. simply click on any video to open it up and you can access extend from there. you can also right click on the player to save any frame as an asset

[charred tendon]

I see, thanks! So on the clip level there's no way to use reference files like ingredients to video. And it looks like ingredients option was also removed from scene builder

those are really great features, can we put those back in?

[viscid mesa]

you can still use ingredients to video on the main grid! there's no option to extend with ingredients

[south hill]

There should be 2 categories for elements which from uploaded and elements which from generated, the all things board seem messy

[charred tendon]

You're right. I'll play around more with the new interface. thanks for the quick responses!

[quartz juniper]

Hi, I have a question. Since this new Flow update, my videos are coming out wrong. I mean, the effects, consistency, and everything else feel really bad. Is anyone else experiencing this?

[charred tendon]

I just downloaded my first 4K clip so far looks okay

The old interface would note in the main Flow window which clips were used in scene builder. Would be nice to have that show on the new main grid.

Also would be good to have clips created in the clip level show on the main grid as clip level clips. Having these videos noted this way on the main grid helps with organization and being able to download the 4K clips you used

For example, I just created a good video on the clip level and I can only download it as 1080p as it does not show on the main grid which has the 4K option

There also doesn't appear to be a way to get back to a clip level unless you remember which clip you used on the main grid

[cyan thunder]

Is the prompt input box now a fixed size? I remember it used to be able to adapt to different sizes.

[tawdry kayak]

im facing the same problem

[cyan thunder]

Hello, can't I download the generated image directly now? It seems I have to go to a secondary menu.

[fickle thicket]

https://play.google.com/store/apps/details?id=com.prominent.aicatch

i am buying like this type apps anyone have to sell ?

[onyx eagle]

I really hope whoever designed this horrible UI gets fired as soon as possible.

[charred tendon]

Correction I do see clip level clips in the main grid, but there's no option to upscale and download them as 4K (720p is the original size) 😟

4K ideally should be an option for everything. I'll spend the credits lol

[umbral fjord]

THANKS

[turbid lintel]

its great , far better than before LOL

we can drag and drop
we can use tag @
the extend and other tool is simple and productive

this is the way

now release veo 4 already !!

[quasi harbor]

Hi everyone, I’m looking for someone to handle post-production for Veo. The assets are ready, but we need to integrate VO, sound effects, and visual polish. The timeline is tight we’re aiming for a release in 2–3 days. Totally open to AI-assisted workflows provided the final quality meets standards.

[edgy holly]

I like it.

[edgy holly]

Gemini is decent at that, but that's not a function I'm familiar with.

[stuck adder]

Exactly hopefully it’s not as nerfed as seedance 2

[fierce nimbus]

there is a filter top right and an ability to filter in the grid top left

[edgy holly]

Can only use videos to extend, but have to pull a frame to use as a reference.

[fierce nimbus]

What can we improve?

[fierce nimbus]

can you clarify this

[edgy holly]

Yeah, it does seem faster. Scenebuilder is still slow to load, but the rest is faster.

[fierce nimbus]

Working on performance everywhere. Projects with tons of files loading into grid/batch FAST was a primary goal of the interface. Just all around speed/performance/stability

[edgy holly]

A 'Copy to' function would work better, instead of having to go back and forth like that.

[fierce nimbus]

have it on our list. Move To

[edgy holly]

That's the same as before. Can only upscale individual clips yet... unfortunately.

[charred tendon]

This is just an identification feature that was helpful to me. Old interface if you added clips from main window to scene builder, it would identify those clips in the main window as "Remove from scene" IMO would be helpful to have that back as well as any clips created on the clip level identified as such in the new main grid. Helpful to know which clips you actually used and liked and then be able to upscale those to 4K from the main window (altho it looks like clip level does not allow for 4K upscale)

[fierce nimbus]

great one. Will add to list

[stray remnant]

im new here looking for help, i make real estate walkthrough videos for realtors and flow was working perfect before this recent update but now when i add an ingredient the output just looks absolutly nothing like the picture i provided anyone have advice?

[twin yarrow]

There were no image "daily limits" on Whisk, but now there are on Flow.... what gives?

[stray remnant]

this whole new update is absolutly terrible

[twin yarrow]

This is a poor update... what a croc.... at least Whisk was free and had no limits....

[stray remnant]

how do i contact someone about it? i just had boughted more credits yesterday but will need a refund its just not even usable for me at this point never in my life have i seen an update for a software do worse for it

bought*

[surreal hare]

Where function create custom prompt climactic @viscid mesa bro?

[edgy holly]

If you want the video to start exactly like the input image, you need to use keyframes, not ingredients.

[surreal hare]

I don't know google upgrade or downgrade function 😭

[stray remnant]

okay ill give that a try

[stray remnant]

wow okay you just solved my problem @edgy holly thanks

[edgy holly]

happy to help

[fierce nimbus]

Working on making this clearer

[fierce nimbus]

whats not working?

[forest ridge]

But in the new UI, it doesn’t show ‘extend clip as individual clip’ anymore. The extended part is merged with the main clip, so I can’t upscale the extended clip separately.

[charred tendon]

yep I noticed that too, can only get 720p

[stray remnant]

just everything about it, its hard to just reuse the same prompt to try it again i have to jump through hoops

[fierce nimbus]

there is a button on every output to reuse prompt. Its even visible during generation. It includes text and media ingredients brought directly into the prompt box.

You can switch to "Batch" mode in settings to view outputs in a reverse chronological list with outputs that were generated together (batch). This is near identical to original Flow

[dry girder]

@fierce nimbus Is there a reason why extended videos are not allowed to be deleted and downloaded with high resolution anymore?

[fierce nimbus]

that one i want to double check on. See it coming up throughout the day so ill add to list of us to prio.

[dry girder]

thank you

[fierce nimbus]

Keep the feedback coming. Happy to answer any questions I can (outside of cost. Im not the cost person). Feedback helps us improve.

[charred tendon]

Nick, appreciate your time and responses! As I've mentioned ideally we'd have 4K upscale as an option everywhere.
Would also be nice as a future feature to upload audio and have characters lip sync to it

[fierce nimbus]

Your goal in the upload audio is for it to literally use that audio in output (puppeteering) or use the uploaded audio to help guide voice/tone with any prompt (reference)?

[charred tendon]

I was thinking more of the former. Is there a practical use case with the latter using Flow?

[cyan thunder]

Is the prompt input box now a fixed size? I remember it used to be able to adapt to different sizes.

[fierce nimbus]

working on an update. Lots of creators are telling us they would like more space for larger prompts.

[cyan thunder]

Thank you. Now that I'm in the image details page, it seems I can't exit by pressing the ESC key.

[fierce nimbus]

on it

[cyan thunder]

Okay, but I meant to press the ESC key on the keyboard.

[fierce nimbus]

Yup. That should work and its a bug. Im a huge hot key fan

[charred tendon]

Can we still upscale images to 4K with nano banana? I only see 1K

[cyan thunder]

yes, bro. It seems like batch image downloads only offer the 1K original size option.

[charred tendon]

like I'm seeing it on a single image download

[fierce nimbus]

2K and 4K available on Nano Banana Pro created images.

[charred tendon]

whoops thanks!

[fierce nimbus]

No worries will make it more clear in UI and ask about support for NB upscale.

[tacit aspen]

@fierce nimbus is prompt expander gone from this release or coming? just curious. One thing I like is all the old videos load faster in each project.

[fluid abyss]

do you guys find it annoying that any upload reference image now comes up along with the generations as well in the gallery?

[fierce nimbus]

Expanders were out of this release. Currently working on how we make it better, easier, more useful. What would you like to see for "sys intrusctions"?

[raven siren]

Wow! Great improvement the new UI! Thank you! Two simple things i''m missing.... Esc key to exit out the edit view and also in that same view being able to us the arrow keys to go to previous and next image...

[fluid abyss]

the best thing about the update now is though we can finally drag and drop or copy and paste an image. this has been a lifesaver

[tacit aspen]

I loved the expander would take a prompt and give me some additional options, camera [director mode] It often enhanced my workflow.

[fierce nimbus]

all uploads and generations are consisdered equal in the grid. Edit the uplaod with nano banana, drag it into prompt box as a frame or combine with other assets as an ingredient. We have sort filters on left side to view grid by content type. Would it be helpful to have a setting to "not show upload in grid" which would make them only avaialbel in @ ref and asset panel?

[fluid abyss]

thank you so much for replying. i now realised that there is a filter for this.
But yes please having a "not show uploads in grid" checkbox on top could make it easier if its possible

and also maybe giving us an option of naming our uploads during uploading or copy pasting would also help us @ref it easier as well. Im unaware if such option already exists

[fierce nimbus]

we are seeing a lot of folks using flow to edit their images. Allowing them to show up next to generated content is our way of saying "work how you work"

[fierce nimbus]

all assets in the grid (uploaded or generated) have a name. Just right click > rename or click on any asset and tap its name to to rename. Once renamed you can use that name to @ reference in prompt, or to search or find in asset drawer. the name is project wide.

[fluid abyss]

thank you so much. The way the new UI is does make a lot of workflows possible now that I think about it. We could create collections like characters, environments, objects. Rename those images and @ref them and maybe also move the generations to collections like "scene 1", "scene 2" or "approved", "optional" and so on.

Very flexible actually. I thank the team for this much needed update!!

[rugged drum]

Would it be possible to allow us to add collections to prompts? E.g. a collection of reference images for a particular character or art style. Rather than adding every reference individually

[fierce nimbus]

Lets talk about it! Each model has a different allowance of ingredients. Veo 3 allows for 3 ingreidents while Nano Banana allows for 10. Neither can accept video ingredients. So we can make it possible to use collections but explaining why we didnt use certain images or blocking use based on constantly changing criteria felt bad. So using collections to prompt is a goal. I want it too. Just wanted to get this out before solving that.

[burnt pilot]

The new flow is confusing
How can i create images using images now i dont get it

[fierce nimbus]

drag and drop them, @ref them, drag them in from desktop, add from desktop from asset panel or frm + button top right.

[burnt pilot]

So i can use it only on desktop now? Ive been using mobile the whole time

[fierce nimbus]

Flow is avaialble on the web. So you can use it in any browser on any device.

[manic sierra]

@fierce nimbus , I’d like to bring a professional perspective to the recent UI changes, as I rely exclusively on an iPad provided by my employer for my daily workflow.

For those of us using tablets as our primary professional workspace, the new "Grid" logic is a massive step backward. The interface has become cluttered and much smaller, making it very difficult to navigate via touch. Most importantly, the removal of the one-tap prompt reload and the easy access to previous iterations has completely broken my creative process. What used to be a seamless, professional tool now requires a tedious, convoluted mess of clicks.

My question is: Is there an active plan to restore the streamlined workflow for iPad/Touch users, or has the mobile professional experience been deprioritized?

We value the output quality of Nano Banana Pro, but the current UI makes it increasingly difficult to justify using Flow as a reliable tool on professional tablet devices. A clear statement on whether these accessibility issues are being addressed would be greatly appreciated.

[fierce nimbus]

You can resize the grid using the settings button top right (we offer small, mesium, large). We have better support for touch on the list but I dont have a timeline for you.

[manic sierra]

Thank you for the quick reply. I’m already using the "Large" setting, but unfortunately, that only affects the image thumbnails, not the accessibility of the text prompts or the overall workflow.

While I appreciate that touch support is "on the list," the core issue remains the loss of a seamless, professional iteration process. For now, the efficiency we had before is simply gone. I hope the team recognizes that for tablet users, this is more than just a minor layout issue—it’s about the basic usability of the tool.

[burnt pilot]

but now on mobile i have no idea how to import images

[fierce nimbus]

• button in top bar or + button in prompt box followed by upload button. Think of the + as "adding the image to the flow project"

[burnt pilot]

@fierce nimbus check dm

[shut olive]

Чому "повторити генерацію" більше не працює? Чому зникає фото яке було додано в генерацію. Зараз я бачу баг, що загрузка нового фото не підтягується, сторінка не опускається нижче щоб додати фото..
Ок, я знайду час щоб щоразу додавати фото, але воно не опускає сторінку що не дозволяє загрузити фото. Я стою на місці вже годину..

[arctic seal]

How to improve video quality?

How to upscale video???

[surreal hare]

Where function custom prompt helper bro @fierce nimbus

[arctic seal]

The interface has become simply terrible. I can't find the button to improve video quality. Many features have disappeared, and it's unclear how to upload images. It downloads some kind of archives... it's just terrible.

[lilac belfry]

Hello

[iron quest]

Can we only download in 1k now? I can't find 4k

[quartz kraken]

Goodmorning. feauture request: when working with nano banana pro, would be nice to paste in refs and images, without adding them or uploading them manually first

[quartz kraken]

maybe take a minute first to learn?

[quartz kraken]

Oh wait, i can paste and they upload into the media, from there I can drag and drop. Nice!

[quartz kraken]

The re-use prompt button is still there? But i understand that it probably doesnt show up when on tablet? Why are you working on a tablet tho? (From a professional perspective 😉

[quartz kraken]

You can find it when you overlay the result video, then overlay (hover) download, there you find the resolution options

[iron quest]

Sorry, talking about NBP gens

I can see it for the videos, not for the images

I've got Ultra sub

[quartz kraken]

i'm able to select download > 4k upscaled there when creating images with nano banana pro

[iron quest]

I don't seem to have that, I'll show, one min

[quartz kraken]

it only shows during hovering though, with the cursor.

[iron quest]

I only have 1k option, before the UI changed, I had 2k and 4k too

[quartz kraken]

oh that's strange indeed!

[iron quest]

Videos I can uprez fine, it's just NBP images 🙁

[jade zenith]

Hi guys are there new limits on nanobananapro in flow? I have a pro and ultra account and seem to be generating maybe 20/30 images which I want to animate with veo3.1 quality before I suddenly get rate limit. Anyone experiencing the same?

[wind saffron]

uploading ingredient images that are PNG, flow is not keeping the transparency now , this is completely useless for me and I'd expect many others? It seems youre converting the files to jpeg?

[rugged glacier]

I was expecting a major update, Flow simply changed the interface, I was expecting lip-syncing, motion capture, and voice control, but I didn't see any surprises other than bugs.😭

[brave pawn]

@anika.labs.google @rugged glacier

[quartz kraken]

It seems to be impossible to generate with nano banana pro in 2k or 4k. only upscaling. any thoughts?

[rugged glacier]

some buttons are missing,😡

[plain robin]

Worst ever update. I am not able to edit the images as good as i was before

[turbid lintel]

nah , this is the best update

[tawdry burrow]

Hi everyone, has anyone else lost an entire folder of videos? And is anyone else unable to upscale videos to 4K? I'm on the Ultra plan.

[hazy obsidian]

i'm facing this error now for more than 10 mins whats happeneing another bug now?

[tawdry burrow]

The new update is incredibly buggy😭😭

[wind saffron]

@fierce nimbus any PNG (haven't tested other file types) created in Photoshop does not upload correctly into flow the transparency is lost, Ive tried all different file sizes/canvas sizes to no avail. If I use another program to create PNG files (although ive only tried Vector Magic) it works fine. This was NOT happening before the UI update.

[arctic marsh]

the new update very very ugly

[hazy obsidian]

@fierce nimbus 😭 please take it as a report and check i have too much to work now

[thick shadow]

This happened to me too.

[prisma ferry]

it doesn’t show up on a laptop either for about 10% of my images

[upbeat sierra]

Hi all,

We value this community and the passion & feedback on the Flow product. We are listening to all of your feedback and doing our best to make improvements where possible. We appreciate the patience, the entire team is passionate about delivering a great experience for you.

We're doing our best to respond to comments in this chat, #1238621511401734254 is also available for easier issue tracking.

[upbeat sierra]

@jade zenith limits should be exactly the same as before, if you're still having this issue, please file a bug in #1238621511401734254 so we can have someone look into this

[upbeat sierra]

Glad you like this!

Dragging and Dropping ingredients and Pasting from Clipboard are two ways we're hoping it is easier to get your ingredients into Flow!

[upbeat sierra]

@iron quest hmm the 2k and 4k options should be available for you on the Ultra tier. Please file a bug in #1238621511401734254 and we'll look into this!

[upbeat sierra]

@arctic seal You can upscale Videos when Downloading. This action is in the top-right three dot menu on the grid view. Please start a thread in #1238621511401734254 if you're having more trouble and the team will look into this!

[arctic seal]

I found out how to do it on my computer, but I can't do it on my phone.

[upbeat sierra]

I can reproduce the issue. That's no good, we'll work on this, thank you for flagging and appreciate the patience!

[burnt oriole]

the scenebuilder is having a lot of problems:

• I cant re arragne clips
• It's very problematic to trim clips as there are bugs like sometimes it shows the last frame of the clip even if it's trimmed or like the playbar doesn't allign perfectly with the clip's actual time stamp.
• It would be nice if the scenebuilder didn't take so much time to open

[upbeat sierra]

@burnt oriole you should be able to re-arrange clips using the button on the bottom right hand side of Scenebuilder (next to the download button).

[crisp basalt]

also scenebuilder was not doing great on my end, it was super laggy and idk what causes it to behave like that, i was just trying to rearrange the clips

[burnt oriole]

yes but for me it doesn't work

lemme try to do a screen recording real quick

[upbeat sierra]

sorry to hear this, a screenrecording would be great. would you mind creating a thread in #1238621511401734254 want to make sure it doesn't get lost 🙂

[burnt oriole]

sure!

[cobalt junco]

fine here so far :/

[arctic seal]

It would be great if you focused more on the mobile interface. The desktop version feels outdated. Mobile is the future — you can take your phone anywhere, anytime.

[cobalt junco]

uhhhh

[upbeat sierra]

Also, thank you for sharing the thorough feedback here @burnt oriole #1476273456038416394 message

The team reads all of this and finds it very valuable. We're committed to making this product work for you all, so please keep all of the feedback coming!

[pulsar coral]

I imagine those are like the placeholder galleries for when you already make stuff or for when we import them from the old Whisk. But if you want to see your actual new gallery collection it will surely be here now https://labs.google/fx/tools/flow

Create a bunch of images, drag and drop similar ones onto one another and you'll have galleries.

[summer rapids]

@upbeat sierra Do you plan to commit UI fixes as you correct them (daily for exemple) or on more major update (weekly or later) ? Asking for the one that seems to be corrected on the message but not available yet

[crisp basalt]

nah

[little fulcrum]

Me too! I just checked and this is day 2 without service. I pay for the ultra plan and my work depends on it. What's going on?

[pulsar coral]

The drag and drop new functions are very useful now, gotta not be afraid to try them, you can just drag and drop the same image directly from the UI into the uploader to keep making images too.

[burnt oriole]

Thanks a lot for saying that!

I uploaded the list of bugs and the screenrecording in #1238621511401734254

Love testing flow and the progress based on our feedback 🙂

[upbeat sierra]

We are commited to pushing fixes as soon as we can. Different issues have different levels of complexity, so it's hard to give accurate dates on when things will be fixed, but we will push updates continually - not waiting for a 'major update' to improve bugs etc

[upbeat sierra]

@little fulcrum really sorry to hear this. We know many of you rely on Flow for critical work and projects. I'm not familiar with your specific issue, would you mind starting a thread in #1238621511401734254 so we can triage and track and make sure someone looks into this?

[upbeat sierra]

Ah @little fulcrum just saw your open thread in #1238621511401734254 no need to start another

[stray bear]

Hey everyone. Quick question regarding the recent Flow update the image upscale feature has definitely been removed ?!? leaving us with 1K resolution
As someone heavily relying on this to produce an animated series this is a massive bottleneck. Having to jump to external tools just to get a usable resolution for production is really frustrating, especially for Premium users
Does anyone know if this is a permanent change ? or is the image upscale option going to make a return?

[pulsar coral]

tbh the old Whisk Nano Banana 1 was only allowing 1K downloads too, that can still be tested atm. But Nano Banana Pro images can be downloaded in 2K maybe that was what you saw and there's a daily limit for that according to your payed plan, 4K download is for the AI Ultra plan.

I've always had to upscale normal Whisk images tbh.

https://upscayl.org/

This one's incredibly good for the Whisk images, keeps their consistency if you choose Digital art mode, and it's free all the time.

It has a bulk uploader too as not to waste time, image can go to like even 16K tho I can't imagine who'd need that, maybe a giant banner maker or something

Normal is around 2K

[stray bear]

that makes a lot of sense! Thanks for clearing that up MegaBuster I'm actually on the AI Ultra plan so I guess I just need to figure out where they hid the 4K download option in the new UI Really appreciate the Upscayl link too I'll definitely use that in the meantime 🔥🤙🏻

[pulsar coral]

Another thing that helps is to prompt the resolution yourself, it's not exactly gonna obey you if you say "4k" but if you end your prompt with something like

"Image quality: Remastered 4K HD ultrahigh resolution, sharpened, denoised, unblurry, no jpg artifacts."

[upbeat sierra]

@stray bear this should not be the case - image upscaling should be available for paid tiers (same as before). can you please file a bug in #1238621511401734254

[upbeat sierra]

Three dot menu on the right of an image in the Tile View. You should see the options for 2k & 4k when you click "Download"

[pulsar coral]

It will try, "Clean lines" helps too, or whatever else you can find to make it better. Specially pro version can do it

[pulsar coral]

yeah tbh a nice built in upscaler would be nice

[stray bear]

@ummerr Thanks so much for the quick help! I just checked the three dot menu in Tile View and clicked Download but unfortunately it still only gives me the 1K option Since I'm on the AI Ultra plan it definitely seems like a bug on my end I'm heading over to labs bugs feedback to file a report now
Really appreciate your assistance!😊

[upbeat sierra]

thanks for double checking! sorry for the bug, we will try to fix asap

[shell sorrel]

I do like the idea of using it more to edit images. The page announcing the new Flow says "a redesigned interface that brings image generation to the forefront." Isn't it really time for image gen in Flow to be able to generate in all the aspect ratios that Nano Banana Pro is capable of doing, especially 3:4 and 4:3?

I get that VEO itself is only landscape or portrait, but Ingredients to Video doesn't even require the images to be in landscape or portrait, and plenty of the images being created aren't even intended to be turned into videos.

[brave pawn]

@upbeat sierra

[upbeat sierra]

great feedback - what additional aspect ratios would be the most valuable to you?

[pulsar coral]

tbh some Nano Banana pro image generators do already obey you if you end prompts with something like "Aspect ratio is 11:21" but Flow's doesn't. But I haven't seen normal Nano Banana 1 doing that unless it's taking the same image ratio from the original image, and flow's can't use the original ratio either, it's locked at 9:16 and 16:9. So yeah.

Flow has the square ratio hidden here if you click on the image tho, and it does work for Nano Banana 1, you can crop the aspect ratio to what you want and prompt the image again. Helps to have used a large image from starters tho, but you can crop it whatever other image.

If you choose the "Freeform" thing. It's a little overcomplicated but it does exist.

Just upload the image and click on it instead of directly prompting on it and you'll see those, pick square if you want it or freefrom edit it then you can prompt.

These'd be normal default aspect ratios, including "Auto" that'd help

[river rapids]

@upbeat sierra I have the same error. I can't download 4K images even though I have an Ultra subscription

[burnt oriole]

i think it's locked at 9:16 and 16:9 is because veo doesn't support image inputs which are of any other aspect ratios

[viscid mesa]

can you share what model you used to create the original image? 2k and 4k are only availble for images created with Nano Banana Pro

[pulsar coral]

Most people would prefer 9:16 anyway, it gives a close resolution to character detail. But yeah let's say you upload images to art sites, doesn't look good there from afar at 9:16. Ideal is around 16:21 or so. It'd help to allow it to be prompted manually.

[river rapids]

ok many sorry. i was with Nano Banana not pro 😂

many sorry

[burnt oriole]

nano banana 2 is already rolled out on gemini app foreveryone. will we have a later rollout for flow?

[tawdry burrow]

I can’t download videos that I upscaled to 4K. The “Download” button appears, but when I click it, it disappears and I can’t upscale the clip again. I’m on the Ultra plan.😭

[pulsar coral]

It does have Nano 2 already, it just gives few tries daily, I think like 10 for free accounts, 20-30 or so for the cheapest $5 tier and so on.

You gotta select the model before making the image.

[burnt oriole]

are you confusing nano banana 2 with nano banana pro?

[pulsar coral]

Yeah I call them the same for some reason, Gemini videos on youtube did.

[burnt oriole]

nano banana 2 is different. it was just rolled out today

[glossy sparrow]

🙏

[pulsar coral]

idk Nano 1 and Nano pro/2 were called around the same during rollout.

[pulsar coral]

Also I've been trying the Gemini Nano Pro and the Flow Nano Pro and they feel the same

[vale zephyr]

Well its because they are the Same..

Nano Banana Pro and Nano Banana 2 are 2 different models

[burnt oriole]

i think google only called it pro lol. so much confusion with their naming. nb 2 is actually 3.1 flash image gen. nb1 was 2.5 flash image gen and nbp is 3.0 pro image gen

[pulsar coral]

If you've payed at Gemini iit will transfer to your Flow account, which is neat since now you get some more daily shots.

[glossy sparrow]

hope nb2 will be available in flow asap

[quick rapids]

hope they don't replace NBP like they did on gemini app, its totally gone there, no way to access it

[pulsar coral]

Oh wait no, the Gemini Nano Pro and the Flow Nano pro don't exactly feel the same now that I remember, the Flow Nano pro is much better since it allows more permissive prompts than Gemini's, it's just a headache to deal with the aspect ratio.

But once you do get an image the art quality is around the same

[burnt oriole]

it's kinda bugging right now. some ppl can access nbp via the thinking model and for some like u and me, it doesn't even appear under any model

[pulsar coral]

But I'm sure Flow's prompts are way more permissive

[burnt oriole]

i think you said the opposite. flow has more restrictions than gemini. flow doesn't accept real people

while gemini can generate all types of images except nsfw

[shell sorrel]

Thanks, I appreciate the aspect ratio question and the risk of UI clutter; but my meta-level pain point is that google has excellent features around things like image generation spread across different products, and yet seems unwilling to simply have all those features together. I pay for Google AI, but whether it's Flow, or the Gemini app, etc, it feels like trying to find a house and never being able to have one that has a kitchen AND a bathroom

NBP is capable of doing at least the following aspect ratios, so for me Flow is only truly valuable if it has the full capability
1:1 9:16 16:9 3:4 4:3 3:2 2:3 5:4 4:5 21:9 9:21

[quick rapids]

problem is whether or not veo 3.1 has those capabilities right, since flow integration was made to feed NBP into veo 3.1. it makes sense to have only aspect ratios supported by veo 3.1 on this platform

[shell sorrel]

Oh, that's kinda useful, thanks. It's a shame that people have to hunt around for partial workarounds to try and mimic a capability the model already has. And it's unfortunate most users won't even happen to see helpful comments like this about the workflow

[coarse bane]

how can i see my saved frames in flow ?

after the update

[edgy holly]

I almost never run into filters, but I do find prompt interpretation in Gemini to be better.

[glossy sparrow]

@viscid mesa dont u have any info bout nano banana 2 availability in flow?

[viscid mesa]

👀 🍌 🍌

[shell sorrel]

I think that may have made more sense when Flow was first started, but the official announcement yesterday said "bringing image and video creation under one single roof", as they are subsuming Whisk and ImageFx (both of which can do many different image gen aspect ratios) into Flow
I also haven't had any issues getting Veo Ingredients to Video to work with any aspect ratio images, it still works fine when the landscape or portrait video is generated

[shell sorrel]

2 eyeballs/2 bananas/t-minus 2 hours?

[fallen citrus]

Is anyone else having problems animating video? I get error everytime

@viscid mesa Hi! Any idea why I keep getting Error everytime I submit a video generation?

[viscid mesa]

our team is looking into this issue right now!

[coarse bane]

@viscid mesa how can i see my prvious saved frames in projects ?

i cant find any of my saved frames

[umbral jasper]

yea im video erros everytime now and i thought it was my prompt

[viscid mesa]

there is a project called "Ingredient Archive" where you can view and copy over your previously created / saved images. we know this a bit inconvenient so we are looking into way to improve this, but your data is still all there!

[tardy onyx]

too small

[unborn bolt]

Does flow dont let you download images in 2k-4k anymore?

[turbid umbra]

is this rolled out? still cant find it

[coarse bane]

i still cant find my saved frames

[turbid umbra]

@edgy holly yo mate are you able to find the Ingredient Archive?

[edgy holly]

Yeah, it's at the top of the main projects page.

[wide shard]

hey just wondering am I only one who think nano banana pro model has been updated and not providing the better results like before

also previous UI that has recent uploads section was way better and quickly accessible

[shy void]

Could someone help me and tell me what my mistake is in this video generation? The flow always modifies my image, even though I've asked it not to change the scene, to follow the photo as a reference for the animation! Is this because I'm using Veo 3.1 fast?

[strange forum]

Did flow get stricter on the safety filters?

[viscid mesa]

this is still accessible on the left side! let us know if it works for you!

[chrome imp]

I think it never got really free, flow filters are too restrictive.

[strange forum]

Normally I can make same image on Gemini as on flow but now I can’t..

[coarse bane]

do you know where can i find my previous frames ?

[edgy holly]

Open the settings and take a screenshot of what you've chosen.

[edgy holly]

With each major update, the filters get better at blocking things.

[fair gull]

How do I add assets I've already uploaded or created in Flow? They don't appear in the asset menu.

[upbeat sierra]

there is a project called "Ingredient Archive" where you can view and copy over your previously created / saved images. you may need to scroll through your projects to find it, apologies for the inconvenience but your assets should be there

[fluid abyss]

when is nanobanana 2 rolling out on flow?

[edgy holly]

Change it from Igredients to Frames

[elder mason]

Images 0 credits since when?

[shy void]

Sorry bro, I deleted the screenshot... I'll change it

[edgy holly]

I saw it. Switching it to Frames will probably solve your issue.

[turbid umbra]

how do I remove an extension if I dont like it?

It seems that if I extend a video its one and done, cant remove the extension after that.

[shy void]

Do I need to select the same image for both?

[elder mason]

?

Not sure it’s a bug?

[open goblet]

Anyone else having an issue when clicking on an image to enlarge it, and pressing the "Arrow" to go back - then it chooses a totally random place in the folder? Instead of returning to where the exact image was located (desktop)

[fair gull]

Thanks. I think I understand the workflow now. It may be a good idea to make the Ingrediant Archive a fixed section so it doesn't get lost among your other projects.

[turbid umbra]

did you find it? I cannot find it

[edgy holly]

Frames means your video will start with that exact image. Ingredients means it'll get mixed up with your prompt and any other image inputs. Give it what you want.