#flow-general・🌊
1 messages · Page 6 of 1
I
VideoFx
Will videoFX be coming out of maintenance mode this week?
Probably tomorrow or Wednesday, they said a week
Auto SFX coming to Veo generations?🔥
Nice, seems like this would do what Sora’s storyboard is “supposed” to do
Very weird but seems like I've regained access to VideoFX but the older interface.
Just a heads up to the staff
same here
no, the bottom says veo 2
oh ok fair
😉
Did they fix it.......
DAMMIT!
Yeah the site works
Its just...... old school
So im guessing they either intended this or its a sign that maintenance is almost done
T2I2V doesnt work
Yeah nothing works
So its still undergoing maintenance
should be done by tomorrow or Wednesday
Its incredible how fast this is evolving.
Getting ready for whats next....!
Tomorrow’s gonna be a big, exciting day for all of us 
Hopefully
Hopefully its actually fixed
Oh...its actually down now
The site was up for like 5 mins
Interesting video from behind 🧐
google illuminate video oreview sacrifice a sheep
It's still down 
Don't spam please
We seen ur message the first time
<@&1366831844502274180> can we ban this guy please?
If your asking for prompts please post it in #flow-tips-n-tricks・✏️
<@&1366831844502274180> just ban him he's not listening
Dude hes gonna keep entering this message
Ffs
Thank you for alerting us, I apologize for his continued spam as we were dealing with the user
I seen that lol
You were deleting them but he keeps reposting it
Just needs banned or timed out or something
Flow out!
we're live??
Never mind reverted
For some reason it's not working for me...
its not ready
200 credits per video generation is pretty expensive.
Thats insane
You have to subscribe
You get a 1000 free credits and than after that you pay 21 dollars a month
21 dollars a month is only for VEO 2 not Veo 3
Yes just saw that!
125 a month
I have Google One. Is that useless now?
Hey where do you see 200 credits for a generation?
#videofx
as I understand google one will become ai pro
Is flow now videofx?
Yup, super bummed out
i mean ill still stick to whatever we have right now c: Im still looking forward to try the whole scene builder
I am subscribed to google ai one, but I can’t generate on flow, why is this?
same
We don’t have anything free anymore
Hellooo, I suggest changing the wording of this to at least specify that you need to be subscribed to the $250/month AI Premium subscription plan to even touch Veo 3 - I was under the assumption I could access it as I am on the Google One AI Pro plan 💔
yeah but it's not even allowing us to upgrade to the required plan. Can anyone find the link?
Too many restrictions on imagen4 gonna stick with chatgpt
for the free credits just hit the subscribe button, for the veo 3 i have no idea
I seemed to be not a part of the CPP anymore is this changed for everyone that was?
Yeah same
I agree, super bummed
seems like fast is 10, full is 100, for veo 2
this is a game changer lol
Yeah I saw the credits, I'm talking about the subscription specifically. I'm unable to upgrade in order to use VEO 3
throwback
Google is seriously nuts - $250 a month for a subscription? That sounds like a lot, they should at least make it a yearly subscription, greedy bastards
Sora charges 200 per month
Could you please confirm this for us?
the server costs for video are INSANSE folks
can we choose only VEO 3? not all want the big package of stuff
so far only can do 2 vids and the 3rd gets stuck at 99%
We had generous access to veo2 for free [thanks]. I don’t mind paying now, we just need to understand the pricing mechanism.
i am in love with flow
Yes, same here. The concept of flow is a complete dream for filmmakers.
Are you able to use it?
yep
I can access it but every time I try to generate, it blocks me
the load on these must be crazy rn
I believe you’re on the Pro plan [veo2]? Veo3 also doesn’t mention how many credits are used per gen.
Faq: Note that Veo 3 is not yet compatible with first + last frame, camera controls, extend or Ingredients to Video capabilities. We will default you back to Veo 2 models when you want to use these features.
lol this is just wild, make scenes and stitch together is so easy
first 30s vid done, i should go back to my day job 😿
Not to mentioned Veo3 isn't even unlimited at this price point 😭
where does it say its not unlimited ?
12,500 AI credits
Monthly credits across Flow and Whisk
how many videos in total out of these credits? ?
10 per video
oh if i was part of the early testers for veo2 do i get any free gens or is it just with a sub now>
You have to subscribe
ah rip
it's just sub now
You get limited credits even on the top plan and Veo 3 costs credits to use
my account is unable to subscribe :/ that sucks
i guess i gotta move to something else, veo was fun
Members of the Google AI Pro and Google AI Ultra plans now have AI Credits that can be used across Flow and Whisk. Learn more about Flow.
if going by this that's like 83 veo3 gens against 1250 veo 2 fast mode gens
dang
all these credits wasted on failed video's
How many Credits do you get on a free plan?
.
Not quite. veo used to deliver a success rate of about 99%.
83 video ... what a bad joke
you should see my library
I’d say "Ultra" is for filmmakers and industry users, which makes sense, and the cost is minimal. "Pro" is for ai creatives, which is reasonable.
for the free 3 month of flow what if Im still using the 3 month from testing whisk?
o veo3 is for the $125/mo plan
It already apart of my crowdfunding list of goodies to get if my crowdfunding is successful so we can finish post production
How is 10 videos per month for creatives?
To make matters worse, its more strict now 😭
100 on fast was what I used all the time.
"Ultra" is for creatives and "Pro" is for Enthusiast or hobbyists
10 videos come on
has anyone been able to sign up using a workspace email?
I don't have access atm in my country for flow but is veo 2 default for any user rn ?
Yes
that's decent then
Could someone with VEO3
access help me test a prompt?
When will Flow support uploading my own images?
Something get announced? 🌊
When it launched
does Flow not work with Google Workspace Accounts?
ended up subscribing with another but found out veo3 is only for the ultra sub
which is fine if you just reroll a few times but if its super expensive it makes being creative anbd experimental kinda risky
Google renamed videofx to flow, but didn't upgrade the model from veo 2 to 3.
If you're on pro, you'll get only 100 videos videos a month
Videofx + veo were really impressive despite all the app bugs. it's so sad to see them cripple it and alienate users.
Remind me to never love or trust Google products.
Google cooked
Congratulations to Google for beating OpenAI on price.
Yeah umm....
Im just going to leave this server now. This is literally expensive. WAY too expensive. I won't be able to do generations anymore because credits go consuming like hotcakes. 😦
if it was balanced out so we dont have to pay for failed videos , by only being able to download the one you approved, mistakes get expensive.
True
It’s so expensive—are that many people still using it?
first veo 2 extensions, hmmm
I don't mind paying the sub costs, it makes sense. Especially going all the way to 100 Veo2 fast videos on the $20 end. I just wish the Ultra tier at least gave something akin to "Unlimited Relaxed" after credits are spent. That's a bit of a bummer. Imagining running out of credits after spending $250 and having to buy more still.
anyone else getting this error?
Ah well. I'm very grateful I had free access to a leading video model for a few month. I hadn't realised it was ending but it was great while it lasted. The price does seem high for Veo3 but I understand Google invested tens of millions training the model and generating the videos is also very costly.
and unkown amount of money maintaining the hardware and keeping it running.
Extend is available in the $20/mo plan 🙂
@hardy halo Is it available in Canada?
Failed videos get refunded! Let us know if you see otherwise 🙂
Currently US only 😦 but working on more countries soon!
did anyone get that email ?
thank you for the email btw ❤️ i cant seems to claim it on my main account, how do i get help for this?
I'll be on for a bit! Send me your Veo 3 prompts 😄
just ran out of credits 10 mins after subscribing :/ veo is a rich mans tool now
@hardy halo any insights on access for workspace accounts yet? i cant redeem the 3 months 😦
i think that was 12 prompts? most on fast mode.
Nope
Just waiting for an in-depth Veo 3 video demo! Really curious about this dialogue feature and Veo 3 in general before I pay into this higher tier.
thankfully it was on the free trial but still
Can you explain the issue you're seeing
when you click the link in the email that says you get 3 free months if you're a pro subscriber already, you get an error page that says your account is not elegible for the offer
This!!
As part of your Google AI Pro subscription, you will receive access to Flow. Click the link below to redeem your offer:
https://one.google.com/offer/...
Already a Google AI Pro Subscriber? You still get 3 months free!
If you're already a Google AI Pro subscriber and you click the link to redeem the offer, your current subscription will be replaced with this 3-month free trial, and any remaining time on your original plan will be refunded. After the 3-month trial period ends, your subscription will automatically continue at the regular Google AI Pro monthly fee unless you cancel before the end of the trial.
the email in question
O- I think you’re not supposed to share that
it's the public email
Exciting news! Today we announced that Flow (formerly VideoFX) is available to Google AI Pro and Ultra subscribers in the US. As a special thank you for being part of our early access community, we're offering 3 months of Google AI Pro for free!
the code there is unique to you
where do i do that at, Im having a hard time getting a desert sand strom working with the actor in it. never makes a sand storm.
To be honest the director and I have been stuck on it using VFX green screen and been trying a lot in Whisk Veo2 and now Flow.
Soon the Director and I was going to ask if anyone in the server wants to help with the 2 scenes, of the movie we are stuck on.
💯 ❤️
Public?
and it doesn't work so doesn't matter
I just came here to say this as well - when I clicked the link from the email, it says my acount is not eligible for the offer
Well, it’s probably too late now someone else might’ve already grabbed it so
yeah. there are 2 emails going on, one to most of the people in the labs test and one to a select few. this is the one to most
makes since.. been a long day
the email is regarding veo 3 or ?
own't do them any good, i had clicked it twice already as it was
.. that’s on you then
yup. i'm not bothered, i've been paying for pro for a long time
veo3 prompt: The pirate red panda beats up the crocodile. 
C’mon, you gotta throw in dialogue, SFX, and music, go all out with the tool!😼
So veo3 is only for the ultra plan
Used the exact prompt, it added the SFX for me 🙂
what are the 2 different emails?
today is christmas morning
@main saddle Could we extend the beta testers' offer to the Ultra plan [Veo3]? That would make sense.
i dont know which toy to play with first lol
Stitch is wild
pro i think?
the Dev side is crazy lol
Can't buy Flow in Brazil
Cause brazil is not a real country
This video is much better than Veo2(fast).Thank you very much.

Amazing!!!!
I've received the email with the fantastic offer of three months free access to Google AI Pro. Regardless of being based in the UK, I was able to access the early version. I was wondering if early testers like me will also get access to Flow? Currently, I can only access Veo 2 on the Google AI Pro account. @hardy halo
That's fire
Any more veo 3 examples? 👀
250 $ for 83 veo-3 videos is "only a bit expensive" for ya ?
You can get more storage if you’re invited to try Google One. Based on the plan, you can also enjoy extra member benefits and share your membership with up to 5 family members. Trials memberships are
i got damaged when i saw that image with the credits and prices ... ma bad :))
button was not a good name to pick for today lol, my phone is melting
?
Same shows for me. And its gemini pro
Yup, I agree. It's for Gemini Ultra only now as an early access. It showed be more widely available in the coming months/year. Did you upgrade to the ultra?

So is free Veo2 via Videofx Beta gone forever?
I do want to try out Veo 3 but I am not able, nor would I be currently willing, to spring for the ultra package simply to use it.
Especially with how limited Veo3 is even at this price
Yeah same haha. Just gotta wait a few months and hopefully it'll be available more broadly. In the meantime, Veo 2 fast mode is enough for what I want too
Great news. Checking to see if Veo3 with audio is available. Thanks
region lock?
Was anyone able to pay for Ultra from outside the US? doesn't work even with a VPN.
Don't think so. I have a VPN and can't even generate them lol. Hopefully it goes out more region wide soon
wouldn't count on that. VideoFX was geoblocked for more than a year.
Is the Ultra tier able to currently upload images for I2v? Or is it just not available at all yet? I'm just seeing "Dropping soon" on the upload button.
Yeah. You can only hope, not much else you can do unfortunately
holy ******* FLOW IS AMAZING
Made a music video with a song I wrote the lyrics for and generated with Suno....unfortunately...NSFW like most of my comedy (the video isn't nsfw....but paired with the lyrics....anyway yea...flow is just...wow
Very unfortunate to have no longer use Flow
how does one turn off the auto sub-title burn in on the videos w/ veo 3? can't seem to turn it off anywhere
i havent generated anything
and it says i got 0 credits
How do u add sound to flow?
Does Veo3 let one use a portrait aspect ratio?
I used CapCut and Suno
You should be able to have dialogue and and have it playback in the video somehow tho
Hey any news when its coming to Sweden???
Nvm. It costs extra
OK, how do you get to the site? I have paid and qualified for it
What’s the website for flow?
Oh it’s still in labs? Ok wasn’t sure if they put it on a new site or something
is there API available for Flow?
Hey can I see it? Can you send a link to my inbox, I’m trying to decide if I want to subscribe to get access to to it, trying to decide if it’s worth it
https://Flow.google works as well!
Oh is veo 3 not on that ai one pro thing? I noticed it said I had to upgrade so I guess it ain’t lol
Nope
you mean AI studio on google one?
Yup
it's not there right now
Oh ok, I got that 3 month trial so figured I’d put it to use, what all can flow do? Same thing veo 2 did? Did veo 2 get any updates since ea?
look at the channels in this discord. you'll notice that the videoFX channels have been renamed flow
it's fun, go play with it. just go to labs.google.com and it's right on the home page
Extend videos you say? Ok that at least is something lol, I used it in early access and was wanting more lol
video is going to be expensive at first there is no way around it
there's not even the storage for universal access lol
Gates get adjusted
day 1😁
Flow😂
Yep.
Hi everyone, my name is Thomas Burton.
I've been working in the VFX film and game industry for 17+ years now and have a specialization in previsualization. I am very much interested in 'Flow' and how this can be used in tandem with film/tv development and would love to be more involved with the team behind Flow.
https://www.linkedin.com/in/thomasburton85/
How is everyone finding flow so far?
No. way
Welcome aboard! You might find these samples interesting. It’s only been a few hours, and the community has already produced some great generations.
https://discord.com/channels/1172568727942860810/1294402841628250162
I understand that top-up AI Credits are valid for 12 months from time of purchase, but does anyone know if unused ultra AI credits from the monthly subscription roll over if not used within a month?
Most likely no
Ah, yeah, just found it in the docs, "Unused AI Credits don’t roll over to the next billing cycle."
@winged plinth yeah I had a browse and am very impressed! The company I'm currently at in Canada is very interested in AI and it's future implementations especially for previs here the name of the game is speed. For now I'm just exploring off my own back but eventually it could lead to more. I see credit limit mentioned alot here are people finding it abit tight at the moment?
has anyone figured out how to guarantee / frequently produce accompanying sound (especially voiced dialogue)?
Quoting doesn't seem to do it, nor describing the sounds in the prompt.
Yes, kind of still adjusting. The veo3 [flow] can be a bit heavy for solo creators, but like I said, industry users, especially filmmakers, should be able to budget for it. Worth it, in my opinion.
Does anyone know how much the promo for early access users is?
You have got to be kidding me
Then why did u link me it in the email bro 🤣
XD
Jesus
Just my damn luck
Seen @torpid cedar latest outputs? Spot on. I’d call them superb.
Free Pro package for 3 months.
@winged plinth wow that lipsync feature is amazing! Very impressive!!! Have seen fully release shows on netflix with worse lipsync (gundam requiem I'm looking at you)
Exactly! It's insane!
Man I can't wait this to come to Canada so I can test this too.
I've been enjoying taking still frames from existing trailers and trying to make better versions of shots with AI. I took that shouting shot from Predator Badlands and it was very easy
Haha, same! flim.ai
I think a good tip to improve all AI outputs it so make sure to describe the camera movement, don't leave it to chance! Adds a really good level of detail to it!
you are the director 😉
Guys I’m sure this has been asked 100 times today lol but how do I get access to Veo 3? Subscribe to ultra? Where?
Thats awesome, thank you
I'll probably just end up buying the ultra plan
@winged plinth Is there anyway to get more involved with flow and the team making it? I'd love to help contribute what I've learnt over the years towards this project.
It does that often
Anyone got this error right after prompts submitted? No video generated but my AI credits have been comsumed.
what I just did on my phone is insane lol, I

@torpid cedar how do we get started with all this? Is their a tester program we can sign up to?
theres no tester, just need to subscribe to google one ai plan
@crisp basalt thanks! Good to know. Am super curious. People in the industry are nervous about this but I see it as a tool to help us do more!
Google Lab's AI video generation tool, powered by Google DeepMind’s latest model: Veo 2
Thank you. Am excited to see what everyone creates!
wait is this coming to flow
whoever did the new design for the veo deepmind page, you rock
https://deepmind.google/models/veo/
I'm so confused. I got the email to get my free 3 month trial sub to Gemini Pro. I was already a Gemini Pro subscriber but I read they will refund the money I already paid this month and then give me 3 free months before they start charging me $20/month. What I'm totally confused about is how do I get access to "Flow"? I clicked on a link that said it was not available to me? Also it seems like the Veo2 in Gemini still doesn't have image prompts yet. Does Flow Veo2 have image prompts like VideoFX did? And is the Gemini Veo2 still throttled at just 5 vids a day?
https://labs.google/fx/tools/flow/ @flint summit
Google Lab's AI video generation tool, powered by Google DeepMind’s latest model: Veo 2
as for now , image to video (using user uploaded image) is disabled
It's in the ultra.
Thanks for the Flow link! That works. Looks kind of like the old VideoFX UI. Now how many Flow Vids can I create before I run out of credits? Oh wait I see there's a chart above. I'll have to read that. lol
Ah, ok. Thought you mean veo3.
I'm still trying to figure out if Veo2 is improved at all? Any new features?
I'd suggest reach out to @main saddle and @hardy halo
yep!
Reference to Video, Camera Control
and start image and end end frame image
@winged plinth thanks so much for the suggestion! I'd love to see what I can bring to the table this is such an exciting time to be an artist!!!
What about audio with people talking? Is that Veo3? Or will Veo2 do that?
yep thats only for veo 3
what about extending? Can you do that yet?
yeah you can !
on your project page , you can click on "add to scene" and it will take you to scenebuilder tool
In veo 2 you can extend?
or just click here too
is audio currently available on veo 3? I can't seem to get flow to generate audio. I'm on ultra, select veo 3.
It syncs automatically with your prompt. Like "rain falling on a metel top" it'll illustrate the sounds with the prompt accordingly
What's a way to prompt voiceover. I tried "Says XYZ etc" but doesn't generate
What you mean? Like a person speaking?
Ya
Prompt: A medium shot frames an old sailor, his knitted blue sailor hat casting a shadow over his eyes, a thick grey beard obscuring his chin. He holds his pipe in one hand, gesturing with it towards the churning, grey sea beyond the ship's railing. "This ocean, it's a force, a wild, untamed might. And she commands your awe, with every breaking light"
Prompt:
A wise old owl and a nervous badger sit on a moonlit forest path. "They left behind a...a 'ball' today. It bounced higher than I can jump.” the badger stammered, trying to comprehend it. “What manner of magic is that?" the owl hooted thoughtfully. Audio: Owl hooting, badger's nervous chitters, rustling leaves, crickets.
A wise old owl flies away out of the frame and a nervous young badger runs in a different direction out of the frame. In the background, you can see a squirrel hurrying past making noise of rustling dried autumn leaves as it goes. Audio: birdsong, loud and leaves rustling, and the sound of intermittent pleasant sounds buzzing, twigs snapping underfoot, and the sounds of squirrels scurrying through the dried autumn leaves. The sound of an owl hooting in the distance, badger's nervous chitters, rustling leaves, crickets, sounds that are full of innocent curiosity.```
I tried to get her to say your name too. But it's on the cake lol
Prompt :
A medium shot, historical adventure setting: Warm lamplight illuminates a cartographer in a cluttered study, poring over an ancient, sprawling map spread across a large table. Cartographer: "According to this old sea chart, the lost island isn't myth! We must prepare an expedition immediately!"
Thanks! these are awesome
okay this definitely convinced me to get veo 3 as soon as possible lmfao
Prompt :
A detective interrogates a nervous-looking rubber duck. "Where were you on the night of the bubble bath?!" he quacks. Audio: Detective's stern quack, nervous squeaks from rubber duck.
i didnt expect a very long length prompt for this one omg https://deepmind.google/api/blob/website/media/veo__veo-3__us_wm_the-scene-explodes-with-the-raw-visceral-and-unpredictable-energ.mp4
Prompt :
The scene explodes with the raw, visceral, and unpredictable energy of a hardcore off-road rally, captured with a dynamic, almost found-footage or embedded sports documentary aesthetic. The camera is often shaky, seemingly mounted inside one of the vehicles or held by a daring spectator very close to the action, frequently splattered with mud or water, catching unintentional lens flares from the natural, often harsh, sunlight filtering through trees or reflecting off wet surfaces. We are immersed in a challenging, untamed natural environment – perhaps a dense, muddy forest trail, a treacherous rocky incline littered with loose scree, or a series_of shallow, fast-flowing river crossings. Several heavily modified, entirely unidentifiable, and unbranded off-road vehicles are engaged in a frenetic, no-holds-barred race. These are not showroom models; they are custom-built, rugged machines – open-wheeled buggies with exposed engines and prominent roll cages, heavily armored pickup trucks with oversized, knobby tires and snorkel exhausts, their original forms and manufacturers completely obscured by extreme modifications, layers of caked-on mud, and a general air of brutal functionality. The dominant sounds are the deafening, guttural roar of powerful, untamed engines, the whine of transmissions, the percussive impact of suspension bottoming out, and the constant spray of mud and water. Within an 8-second sequence, one of the lead vehicles, a low-slung, open-cockpit buggy so caked in thick, brown mud that its original color is a mystery, approaches a wide, shallow river crossing at incredible speed. Without the slightest hesitation, its unseen driver powers straight into the water. The impact sends an enormous, almost solid, opaque sheet of muddy water, mixed with stones and debris from the riverbed, spectacularly high into the air, completely engulfing the small buggy for a terrifying moment, obscuring it from view as if it has been swallowed by the river itself. Right on its tail, a pursuing, equally mud-encrusted, custom-built truck – a hulking, high-clearance beast with a heavily reinforced external roll cage and no discernible badging – arrives at the river crossing just as this massive wall of airborne water reaches its peak. Instead of slowing or attempting to find a clearer path, the truck's driver, with unwavering aggression, plunges directly into and through this opaque, turbulent curtain of muddy spray at full throttle. A split second later, the truck bursts out from the other side of the deluge, water cascading from its roof and chassis, its oversized windshield wipers struggling frantically to clear the torrent of muddy water obscuring the driver's vision. It lands heavily on the far bank, suspension groaning, but still in hot pursuit of the now-reappearing buggy. This thrilling, messy, and visually spectacular sequence of one vehicle creating a massive environmental obstacle and the next immediately conquering it through sheer force, forms the core, immersive, attention-grabbing event of the 8-second sequence. The race continues with undiminished ferocity, the natural terrain itself an active participant in the conflict
the rest yall can check out the deepmind veo website, its very informative and useful for guidance https://deepmind.google/models/veo/
Monster truck!!
Anyone able to use this in UK yet?
Can someone w veo3 ask it to create a video of a vehicle reversing and tell me if it does it. Veo2 was not grounded in physics
how do i add sound? im confused
How do you make a short film using Veo 3?
By using the scenebuilder tool
This is actually so cool that they included these benchmarks
Transparency for the win 
I get too many errors, inculding sense video can't extent, can't and frame as asset, .....
Can you generate 9:16 videos in veo 3? Haven’t seen anyone do that yet.
Got my answer. Yeah it’s a no go for me then because it can’t do 9:16 ratio
a women using watecolor paint set
is upscaling working in flow?
i ran my first test with veo 3 and it said upscaling failed
how can i use veo 3
Pay
I am a Pro subscriber and I always get this error when try to generate, any one could help?
I want this to allow me to upload my own images.

Upload to your Google drive you can use them that way
Hi guys 👋🏽
Anybody know if using Veo2 with AI Studio is the same than using Veo2 witht Flow?
It's only available in the US for now
I don’t seem to see an option to open google Drive.
I don’t think it works like that, those images in the list are usually ones the user generated and/or uploaded back during the early-access phase of VideoFX.
Oh that’s true
Wassap guys, how to get Flow? Amazing tool
Subscribe to Google one ai premium plan
Google Lab's AI video generation tool, powered by Google DeepMind’s latest model: Veo 2
Thank you
Gotta say, the start and end frame feature really blew me away with how well it generated the video 
Hello everyone. My name is Sean. I'm the founder of Screamify.com. I've been playing with Flow all morning and it's clear this is the most amazing filmmaking tool to come out since the camera. Screamify is a brand new streaming service so if there are any AGI directors who are making short films in the horror / thriller category I would love to feature them on our platform. My email is sean@screamify.com
I am a google one subscriber but on flow it say i need google ai subscription, when i click on the link, it shows i am already subscribed - so what is it?
I really hope they eventually let Pro users top up credits too.
Are you from US?
yes
If you tap your profile icon at the top right, it should show your AI credits and whether you're on a Pro or Ultra plan, do you see that on your end?
Hi everyone I’m Lobe from the US
Hi, I'm Ben, just signed up for Ultra and I'm trying to use flow. When I try to use scene builder with Veo 3, the sound stops working whenever I try to add more than one clip to the scene. Is this expected behavior?
In scene builder, if i right click the composite video and download it, then the downloaded video has sound as expected. So maybe this is a bug.
💡Google Flow Mini Tip — Rename Your Projects!
Did you know you can rename your project directly from the main page of Flow?
- First, switch to Grid View by clicking this button in the top right.
- Then, hover over the project name (or date) and click the little ✏️ edit icon
@latent elbow
I accidentally generated videos with Veo 2 at least 16 times and was wondering if they nerfed my outputs 😔
Is there any chance to get an access from EU?
already used 37.36% of the credits in basically 1 hour : (
Does the audio only work on text to video?
veo 3 model specifically
Is there a specific way to get audio? I’m about to make my first clip and I don’t want to waste the credits not knowing what I’m doing lol
Make sure you pick Veo 3 model
as for prompting , you can refer to this
A medium shot frames an old sailor, his knitted blue sailor hat casting a shadow over his eyes, a thick grey beard obscuring his chin. He holds his pipe in one hand, gesturing with it towards the churning, grey sea beyond the ship's railing. "This ocean, it's a force, a wild, untamed might. And she commands your awe, with every breaking light"
some more example :
https://deepmind.google/api/blob/website/media/us_wm_a-detective-interrogates-a-nervous-looking-rubber-duck-where-wer_.mp4
A detective interrogates a nervous-looking rubber duck. "Where were you on the night of the bubble bath?!" he quacks. Audio: Detective's stern quack, nervous squeaks from rubber duck.
Ok, I appreciate the help
Ah so far 2 videos waisted it’s not understanding what I’m asking, even though I’m keep it clear and simple
Yeah I need to wait to be able to upload my own image assets
I mean how can I use frame when I can’t even upload images 🤷♂️
I mean like a conversation between two people, it’s not making the people say what I asked it’s mixing it up
Whats the prompt if u don’t mind
I’ll be back home in a second and I’ll post it and the video it made. Had to drop my son off to school
this is how to 'rename' a project:
Interesting
Getting this message, even though I'm already subscribed and in US: You must be a Google AI subscriber to generate. Feel free to look around in the meantime.
I didn't receive any promo code. Who received it?
Check your spam folder
when was Veo 2 first shown
Contact the help section. Or ask a mod for further assistance
Just do search in your Gmail “flow”
How it works?
Do you need credits to use Flow?
Ooooh
Where's generate video button?
Yes
Goddamn, thanks, bro!
I want to turn images into video. Can i do this with Google AI Pro Subscription?
daily?
I'm considering upgrading from Gemini Pro to Ultra, mostly for VEO3, and have some questions, in case anyone can share their experiences.
For people using VEO3, is it randomly adding subtitles to videos, which I've seen some people mention. Is this the case, if what are you finding the rate of random subtitles to be? Based on looking at some early videos that people have shared, it roughly looks like it might be 15%. My concern here is that, for me, such shots are useless, and throwing away a significant fraction of the 83 shot one can generate with the plan is less than ideal. Should I be concerned?
Flow helps with character visual consistency, which is great for moving to longer narrative videos. Is there anyway to get a consistent voice from generation to generation, or is this something that vaguely determined by the model by based on your prompt, and not something you directly control?
Is there anything else you'd suggest I consider about a considerable expense (for me) for hobby purposes?
Hey there! I wanted to try out Flow, but even with a Pro subscription, it says I have 0 credits. Is anyone else experiencing this?
Were you an early tester previously? If so they should have sent you an email with a promotional code.
Nope, I'm not an early tester. Is it a prerequisite to be able to use it with a Pro subscription?
Do I need an US Google account to get access?
Not really besides just purchasing a pro plan entirely new
I'm pro for a month or two already
I have AI Pro sub but unfortunately the Flow does not register it...
How to fix this any1 has the same problem?
Same issue here.
😬 this is something Google labs team need to address ;
Hey folks, same here. I´m already a pro subscriber and wanted to upgrade by selecting the ai ultra subscription, bu got the message, that I´m already a subscriber. Can anyone help?
Probably a bug
i put this under bugs feedback channel pls engage https://discord.com/channels/1172568727942860810/1374753245142192179
Hey Guys, Has anyone had the issue with no videos having audio? Only the first video I ever created had audio - even when I download the video it still does not have it meaning it's not a browser issue.
peopel on twtiiter spreading lies saying you can only make 5 videos a day with google ultra plan lol
Is veo3 limited to five uses per day?
please @fierce valve give me the access :/
why my current gmail is down for using flow?
anyone else getting this error in Flow today?
same bro,
ok, at least i'm not alone lol
is anyone else's flow down
yeah it seems like its down atm
its says i need to subscribe but i already did
Hopefully they turn on the other features we paid for lol
img2vid outside of imagen is needed
why I going to pay something that I can used free
ingredients would be killer
is anyone else having it say thye need to subscribe when they already did
why upload images is not available ...
i am on pro with google one plan
upload images is the best part of videofx
Hello, trying to use veo3, set my prompt but I'm not getting any sound.
you must click on high quality
Ah I see it now. Thank you!
only i get 1000 credits?
is it still down for you
yes bro, because only 1 video is 100 credits thats a lot
it works fine on my end
Yeah its still down for me
won't load any credits. have logged out and tried diff browsers
i notice u dont have the "pro" tag on ur icon
When will flow be available in Poland
It won't load it at all. I have ultra
youre from us right?
yep US
click on your profile icon , do u see how many ai credits u have?
yeah i think it has to do with,, idk cookies related issues ,. dont quote me on this tho, but have u tried clearing ur browser cache
ok since the ai credits work accross whisk and flow, can u try open whisk?
see if it shows credit balance
Yeah, doesn't work in flow either. So its gotta be my whole google account
I mean whisk
won't load credits anywhere
still got my 33tb of storage tho lol
The storage will never be a problem
oof, if u cant even use both tools (whisk and flow) .. i think you should ping one of google staff
A few people above also had this issue it seems. Might be related to upgrading possibly.
BTW, this isnt really a "must" , but if you want to help the staff understand your issue better, open the inspect element/dev tools, go to the network or console tab, and screenshot the exact error message shown. That’ll give them clearer insight into what’s going wrong.
Is there anyone on the Google team that is active here. I'd like to reach out about account access for our team. You can ping me directly too.
ok last resort would be... login to flow on new incognito tab
unfortunately, still the same 😦 gotta be the account
Ok looks like more people are getting same issue as yours, i hope they look into this
mine is still working on my end but i might spoke too soon
thanks for the help anyway 🙂 hopefully they fix it for us excited people
When will flow be available in Poland
after you topple the EU. have you tried a VPN?
Did you find out ?
I HAVE OVER 7000 CREDITS AND I CAN'T USE VEO 3
flow is not working
is it broken for anyone else?
Same here
My first video, any thoughts?
https://youtu.be/TPkXEV4IWXc
Trying out Gemini's AI tool Flow video creator.
Subscribers 0
the sudden jump to bus scene caught me offguard haha , but overall it looks good
Yeah, I would have to prompt it with more detail on how to get on the bus. But asking the duck to get on a bus without any detail it was not bad as a transition.
how do you use flow is it only avaible in us ?
Google search says limited rollout. Do you have AI Pro subscription?
no i dont, need a gemini subscription ?
You'll need Google AI Pro or Ultra subsciption plan to access....
Good thing is that you can get like a free month or something if you really want to try it.
It does say, limited rollout to USA first. Not sure what other countries are next.
ok and is there a limited credit thing ? i got kling gpt and midjourney subscription and wanna try google tools
Has anyone tried accessing veo 3 with a VPN and if thats possible?
Any guesses as to when we'll be able to upload frames for image to video? @fierce valve
Veo 3 is with the ultra account only? I can't afford it. 🥲
Even with a VPN, Flow still checks if your account’s Google One AI plan was purchased from a US region.
Before you ask: yes, everything is AI here. The video and sound both coming from a single text prompt using #Veo3 by @GoogleDeepMind .Whoever is cooking the model, let him cook! Congrats @Totemko and the team for the Google I/O live stream and the new Veo site!
just because i want the playback embed , hope u dont mind ❤️
SICKKK
oh god if i show this to my dad , He would not tell it's ai generated
sure, thank you!
Really love how the storyline flows, feels genuinely like an authentic car review video.
Just a quick reminder for everyone:
! before generating anything in Flow, double-check your model quality setting, especially if you're aiming to use Veo 3 with Audio.
Saying this cause I didn’t realize it automatically switched to "quality" mode after I submitted a prompt, instantly burned 300 credits 😭
How detailed were the prompts? Mind blown by the people and cars in the background
How do I opt my workspace into allowing me to use Veo/Flow? I try to sign in and continue but the accept privacy policy button is disabled
noob here..how do i stop the subtitles in my video generation or at least have the subtitles not have so many grammatical errors. using ultra
Does anyone know the ETA until you can upload pictures on to Flow? I subscribed to the AI Pro plan without realizing this wasn't a thing yet.
If it's going to be too long, is there a chance I can cancel my subscription with a refund?
Probably can get a refund.
same
ye kinda signed up for that purpose too lol. got some concert visuals to create by June 😩
can someone tell me how to generate vids without the grammatical errors in the subtitles
It's hard to tell as it is but if you added noise and downgraded the video I don't think I would be able to tell that this was generated. those little AI tells would be lost in the noise and lower quality.
I love it!
but man!
I can't help be more concerned about misuse then I was when I flew the first generation sub $500 DJI Phantom 1 and realized you could strap a small bomb on it and fly it anywhere. that was about 15ish years ago. This is the first time it's hitting me that our world is not going to be the same and unlike with the drones where you could easily see how it could be misused. this is next level change because it's one spoke of the AI wheel.
again I love it!
but I can't lie that part of me wants to see if I could join one of those uncontacted privative tribes.
Skynet by the year 2050
anyone can help? I have a pro subscribtion but can't generate. here is what i see: Flow is available exclusively to Google One AI subscribers. Not a subscriber yet? Take a look around.
server issues
im not asking for subtitiles. I even put in my prompt "do not use subtitles" but it still does.
I didn't find answer but I think Veo2 on Flow is better. Always old version on AI Studio.
Always old version on AI Studio. - anywat to have access to it still?
damn those are good. im upset i cant use this right now
i know most of these looks like it was made with veo 3 (mainly because they have audio ) I wish I could use it too. Really hoping they eventually make Veo 3 more accessible or affordable somehow 
I'm willing to spend some credits for those really really wanting to try. Give me prompts for the initial image, then extend the video a few times by additional prompts.
Hey guys! New here - is there a way to get the same avatar/person in the next prompt? I am using VEO 3 and Im trying to get it to work
Oh my God. Work just said they're buying ultra for me.
Anyone have any links for Any of the first movers who are testing out the new video model?
I know we have a training session tomorrow morning, but excited to start working early.
it was shown that at some point you will have reference for characters, scenes, but I think it's not available now.
you could do image2video though, and get the individual shots with character preservation from a different platform
I'm using Veo 2, but the extend button works for the following prompt that extends the scene, instead of generating a new one.
Problem with that is then there is no audio
Where do you work, can I get a job there? 😆
when i do the image to video or ingredients it downgrades me so there isn't any audio
Yeah, no audio with Veo 2 / pro plan.
Hello, is it possible to generate image to video with whisk?
I'm at a school in Hawaii who's doing AI literacy training for their students. 😎 ☀️ 🌊
Hopefully we can host anothet conference or something so everyone can come out and hang out again!
Hey friends. Anyone have concrete info on the maximum seconds in a single Veo 3 with speech clip?
dope, super important to learn these skills early, but also to keep up in today's evolving tech
Also, has anyone figured out how to create audio with ultra / veo 3? My prompt isn't making the character speak.
you can refer here : https://deepmind.google/models/veo/
scroll down a bit and you'll see the prompt examples
Is it possible get flow outside usa?
Prompt: A keyboard whose keys are made of different types of candy. Typing makes sweet, crunchy sounds. Audio: Crunchy, sugary typing sounds, delighted giggles.
Prompt: A woman, classical violinist with intense focus plays a complex, rapid passage from a Vivaldi concerto in an ornate, sunlit baroque hall during a rehearsal. Their bow dances across the strings with virtuosic speed and precision. Audio: Bright, virtuosic violin playing, resonant acoustics of the hall, distant footsteps of crew, conductor's occasional soft count-in (muffled), rustling sheet music.
yes, i have prompt ig let me share: A Pakistani man and women riding horse in Karachi and talking about Karachi traffic in Urdu, it generated this. you might try similar to it
oh even without dialogue line in the prompt . it still generate them speaking
Exactly
Hey how do I use veo 3
@crisp basalt I have been stuck here for hours, what should I do ?
I think Flow is designed to be used specifically on desktop Chromium-based browsers.
So you might need to try open it on your laptop/computer
are you on iOS? I am able to use it on my google pixel 9
Could someone with access to Veo3 generate a video for me testing if Veo3 can render realistic physics in a clip of wine glass? Veo2 couldn’t handle this, the wine flowed through the wineglass
Generate a photorealistic video of a transparent wine glass sitting on a table covered in a white tablecloth. A wine bottle enters the frame and starts being poured and red wine fills the wineglass. The bottle continues pouring even after the wineglass is filled, causing the wineglass to overflow as wine spills over onto the white tablecloth.
Did you try on Veo 2 yet?
It failed, wine flowed through the glass. It surprised me how bad
Do you know anyone here with access to Veo3 who might be willing to test this?
you can purchase it for 0$
for trial ig
lol
That’s so strange. It’s “better” than veo2 but still not understanding physics.
ohhh
can someone generate me a video in veo3 or veo2 to test how it looks and physics? PROMPT: Top-down macro shot of clear ice cubes resting on a smooth surface. One ice cube slowly drops into thick orange syrup, creating a beautiful splash in slow motion. Syrup droplets rise upward, suspended midair. Studio lighting, vibrant colors, crystal-clear textures, cinematic food commercial style.
Generated on flow using Veo 2 for me.
i can not get the friggin thing to include any audio
^
Wait, this was Veo2 right?
oh maybe not it was in flow
Right veo 2, prompted in Flow. And not on the Gemini window.
Yeah I want to see Veo3, I have already tested with 2. Your clip wasn’t as blatant as mine but it still magically permeates through the glass
Veo 2 in the Flow prompt window
I have not had this much fun in a while!

thankkks. not really realistic ;))) I am thinking how to do the shot: the ice cube is falling into caramel syrop and goes under the surface with splashes
did you find a resolution for this? same issue here
is there any difference between two? i mean Flow and Gemini Veo2?
same for me
Haven't tried the same prompt on both yet. You'll probably get different results as each time it'll generate something different.
how i can acess veo 3 it says i cant acess
when will be able to upload images to veo 3?
when will be able to upload images to veo 3?
If I subscribe to the Ultra membership, can I upload my own images to generate videos?
Not yet. Based on other user's experiences shared here.
Probably need more safety checks first. You know people gonna push the limits of what they upload.
this
theres tons of other video gens that allow it.
I heard that with a subscription you can use VEO3 only five times a day. Is that true?

Thank you for that. Apparently there are many ways to invoke audio, none of which have worked for me.
did u make sure that you picked the veo 3 model?
Well, I shouldn't say none. I'm actively working through them
some users reported that they just realized it reverted into using veo 2 model automatically
Huh. No, I didn't. Where do I do that?
OMG
uh oh
Thank you.
Anyone else keep getting annoying subtitles on the dialogue within their clips with audio?
I think the Labs team really needs to acknowledge this issue already, haha.
Make sure to check the model setting every time you generate a new video cuz after each prompt submission, it tends to revert back to Veo 2 quality mode, which seems to be the default
It seems the Google Cloud API also supports VEO3, but I’m not sure if it supports using your own images.
Nahh, I use Redmi note 14 but I just switched my Chrome browser to desktop mode so it is working fine now
Im trying to create an AI video clip but they dont have any sound
I enabled my browser, and still no sound even when i download the video and play it in quicktime
Can VEO2 automatically generate voice-over?
anyone else run into this problem? I just made the account to ttry it out
Amazing. It's working. Thank you.
whats with the subtitles?
i need help
Just a heads up - if a veo 3 generated video doesn't have sound, refresh the page. (Using chrome on android in desktop mode)
Hi, I'm new and stupid. I have a new and stupid question....
How many minutes of footage can I make with VEO 3 in a billing cycle?
I have multiple documentary channels on YT and each video is currently 20-25 minutes long.
Can I use VEO to create the entire vid?
The more I get into this, the more I am stunned. Amazing.
Can you not use real peoples names for AI prompts?
Hi! Can you try restaring your computer and/or making sure your browser is up to date?
Just so I am clear, if I upgrade to Ultra I can then use Veo3 in Flow?
Meghan, am I able to start AI prompts using celebrity names or no?
whenever i try to flow cancels it
@small tide is it possible to generate same consistent characters? 🤔
Hi there! We do have some restrictions on the promo based on the type of account you're using. See more below:
Use a Personal Gmail Account:
The trial must be redeemed using a personal Gmail account (e.g., johndoe@gmail.com).
Accounts associated with Google Workspace, educational institutions (.edu), or corporate domains are not eligible for this specific promotion and will result in an error.
Form of Payment Required:
A valid form of payment is required to initiate the trial. This is a standard process for many trial subscriptions.
If you are concerned about being charged, you can typically cancel the trial immediately after redemption through one.google.com or via the Settings in the Google One app. You would still receive full access to the Google AI Pro features (including Flow) for the entire 3-month trial period if the redemption is successful.
Existing Trials & Stacking:
This free trial period cannot be added to an existing free trial. You can only use one trial at a time, and promotion lengths are not stackable.
If you have an existing trial and attempt to redeem this new 3-month offer, your current trial or subscription plan may be replaced or revoked, and the new trial may not activate as expected if there are conflicts.
Can you share a screenshot of what you're experiencing?
Hmmm, i see
it says it violates the guidelines and i dont know why
Thanks for the info
Hi, I just changed my country on my google account from India to US to use Flow, by creating a new US payment profile and changing by google address to a US address. I don't see any AI credits though. I am also Google AI Pro subscriber.
My google one account's payment uses a US card.
I am also the family manager of my google one family.
<@&1173646484311384114>
time to take the plunge!
Is there an ETA on that?
Anyone playing around with using Veo 3 to create TikTok content? the short video limits actually work pretty well given that format
thinking like UGC type stuff
Hey 👋🏼 how do get hold of this free promo on flow I am in sweden does not look like I can buy ultra already have ai pro
hey, does anyone know why do i get a "Flow is not available in your country yet.", my location is US
Hey @small tide so you're saying I'm getting this message possibly because I'm already on a Google One trial? What do i do to upgrade. everytime i click ont he ultra subscription, it brings me here
is there a way to add an image to the flow video prompt? Like: create video of me walking in a park and show a tree (us the tree in the uploaded image)
- (use the tree I uploaded)
That trial only gives access to Veo 2 and not 3 right? Is there a special offer for those who want to get VEO 3?
The special offer for Ultra is $125/month instead of $250/month for the first three months then it jumps back to $250.
#flow-general・🌊 A highly detailed, cinematic animation of a miniature Chilean dish—such as a plate of pantrucas, pastel de choclo, or a choripán with pebre—being "constructed" by tiny, animated miniature figures dressed as workers. These tiny characters are actively assembling the dish as if it were a construction project:
Some are pouring caldo (broth) from a crane into the pantrucas.
Others are carefully placing slices of Schwencke longaniza onto the plate with precision tools.
One team is spreading pebre using trowels, while another group climbs up mashed potatoes or choclo puree, treating it like scaffolding.
A small forklift carries cilantro leaves, and workers carefully position them as garnish.
The camera slowly orbits the dish in 360-degree movement, capturing the textures of the ingredients from all angles—steaming broth, glossy sausage, rustic wooden plate, and fresh herbs. The lighting is warm and cinematic, with soft shadows and rich textures, enhancing the realism of the scene.
A Schwencke logo or product package is subtly placed in the background or on the "equipment" used by the mini workers, grounding the branding in a playful yet high-quality way.
The animation should feel seamless and magical, drawing the viewer into the handcrafted richness of Chilean cuisine—with a touch of humor and wonder.
Yeah But I can't even click on it, it takes me back to the "you are already subscribed page"
So I got my free three months of Gemini Pro w/Veo2 for being a Beta tester. I was already a Pro subscriber but they say they'll refund my monthly payment and give me three free months. That's nice, thanks. But what happens if I sign up for Ultra at $125/month? Do I basically forfeit my three free months of Pro?
Yeah I know, I just can't access it. I'm willing to pay but it's not letting me
Yeah I'm stuck in Canada for the next 2 weeks and can't access it 🤦🏾♂️, Guess I'll have to wait until I get back
Has anyone had any luck getting it to not produce captions while still producing good voice?
One shortcoming seems to be voice control - I've done about 10 different 2d animation style videos and they all have this upbeat peppy female voice regardless of how i try to prompt away from it.
Can flow take image stimulus uploaded as reference points to include in a video
e.g. person's appearance
Negative, I get the captions every time. Mostly with misspellings
This part sucks. Burning tokens.
A friendly reminder for everyone:
Before generating anything in Flow, make sure to double-check your model quality setting, especially if you're aiming to use Veo 3 with Audio.
is there a limit of video you can generate each day?
So how many video can you generate each month, I heard that some people burnt all their credit very fast
Veo 3 is only for those who have the ultra subscription?
is it worth it?
When is "ingredients to video" dropping. I'll cancel my runway ML in a heart beat when I can use my own images and assets.
im hoping that happens soon, need to produce some stuff by June 😩
Hi, I’m new here. I’ve already subscribed to Google AI Pro, but I still can’t access the Flow TV feature. When I try to generate a video, it doesn’t work and my AI credit still shows 0. Also, when I open https://labs.google/fx/tools/flow, the status just says “You must be a Google AI subscriber to generate. Feel free to look around in the meantime.” with no actual content or access to the tool. Please help.
kindly help @small tide
I think its only available in Usa for now
Does Flow only work now if you subscribe to google AI pro? Is it no longer available for labs testers?
Its no longer in testing phase so
I did 😩
do you have the pro or the ultra subscription?
We had some luck with "no captions, no subtitles, no text" but it was still hit or miss.
I just burned through my 1000 free credits. Its only 10 Veo2 hi-qualities. But I kind of wanted to do that before I subscribed to Ultra. I think I may wait until image prompting works before I subscribe.
@hardy halo thanks a ton for the heads up! ❤️ Hopefully that clears things up and more users get a better understanding now.
Can't upload images... Anyone have any intel on workarounds?
Have the "ultra" membership, BTW.
There isn’t workaround, its temporarily disabled
Argh. Thanks, @crisp basalt
Ok... Here's another question: I've created a character... I just want the camera to push in slowly and have him smile.... But the video keeps rendering of the character talking. (No sound. Sound doesn't seem to be working either.) Any insight into a proper way to get the result I'm looking for?
What did you write for the prompt?
Hmmm... Lemme see if it's still there.
I wrote:
Create a character. The character will be a stereotypical shipping package, with tape and shipping a shipping label, etc. Make it friendly looking. It just smiles, posing for the camera in an establishing shot. It does NOT talk. It ONLY smiles. Make it cinematic and photorealistic. The shipping label simply reads, "Pat". It should be placed on the bottom left of the character
I think the prompt is generally strong. The 'DO NOT talk' part might be tripping up VEO 2. Instead of telling it what not to do, you could try something like: 'The character should express emotion through its posture and eyes, without any open mouth or speech.' or simply let the visual details of the shipping package imply its non-verbal nature
Thanks again, @crisp basalt
You might get better results by framing everything in terms of what you want to see, rather than what you don't 
I'll try... FWIW, I only added the , "do not," after it kept giving me the same moving mouth when I asked for it to just smile.
A photorealistic and cinematic establishing shot of a friendly, smiling character. The character is a stereotypical brown shipping package, meticulously detailed with visible tape and a shipping label. It is posing naturally for the camera. Its only form of expression is a wide, continuous smile; it does not communicate verbally. The shipping label, clearly visible, reads 'Pat' and is positioned on the bottom-left surface of the package. The lighting is soft and even, highlighting the textures of the cardboard and tape
@crisp basalt Fantastic. Thank you!
You’re welcome ! 
Crazy how this works. Sometimes it adds the, "Pat," label, and sometimes it doesn't.
This is a creepy ahh amazon box xD
why is the cardboard box have stretch marks? that's a little funny
I ran your prompt through Gemini 😄 Even asked it to make sure it follows the official Veo 2 prompt guidelines. If youre ever stuck with prompt issues, its a good idea to ask Gemini or any LLM you prefer for help 
Has anyone had luck getting consistent characters to speak across scenes? I used text-to-video on the highest quality setting for the initial clip, then pulled a frame from that and ran it through Frames to Video (also on highest quality). I can load the frame fine, but haven’t been able to get the character to speak consistently.
Ultra
Hmm, I'll DM you and let you decide. It still can't read/speak Prefixes correctly (bummer- for us uni folks) but its majorly improved in terms of clarity.
someone know how many credit a 8sec video consume?
Did anyone try the outpainting feature of Veo 2 yet? I can't get on the user journey to do it
Depends on the Quality (and the model behind it). You can see it after clicking the "Settings"
Cant even afford it 😦
Yes for example generated a landspace video, want to generate upper and lower parts of the video as well, to make it watchable in portrait mode.
It's listed under Veo 2 capabilities here: https://deepmind.google/models/veo/
Its not ready yet
How do you know? Do you know when it will be ready?
#flow-tips-n-tricks・✏️ humen and phone
When Google says ‘dropping soon,’ does it mean in a few weeks or in a few days?
Is this a known issue?
AI Search says Flow supports uploading elements and reference pics for use w Veo. Is this true? If so are pics of people supported? https://youtu.be/9Qs5-_DOkVE
Google I/O AI updates: Veo 3, Imagen 4, Flow, Jules, Stitch, Project Astra #ai #aivideo #veo3 #aitools #ainews
Thanks to our sponsor Abacus AI. Try ChatLLM with DeepAgent here: http://chatllm.abacus.ai/?token=aisearch
Veo3: https://deepmind.google/models/veo/
Imagen4: https://deepmind.google/models/imagen/
Whisk: https://labs.google/fx/tools/...
nudging in hopes someone found a solution 
u should join in if u have chance
where is this pinned?
Here it is. Ingredients to Video. It’s odd because the feature appears to be recent.
same issue here
its only for ultra plan users only
Oh so it is available?
yeah
plz help! I have same issue about sound, and I have checked the veo3 setting every time, but most of generated videos still have no sound.
Flow also supports frames to video which supports first and last frames of a video.
that is also available right now
Google Lab's AI video generation tool, powered by Google DeepMind’s latest model: Veo 2
like literally we're all playing with it atm
Wow
I believe were running into the same issue - seems like its just not capable yet
im trying to get consistent characters to speak across scenes
It looks like I can only use ingredients with veo2 🤔 is that right? When I try to switch it to veo3 it tells me that veo3 doesn't support ingredients or something.
Just a heads-up: the model setting automatically reverts back to Veo Quality mode (Veo 2) every time you submit a prompt, so make sure to double-check before generating again.
But all of this is restricted to a credit limit, right?
yep
Veo 3 only supports Text-to-Video and Frames-to-Video (first frame).
I can sure that before every generation I checked veo3 mark
haha, ah no speech when using frames, that explains another issue I've been having 😅 neat chart
may i see how you wrote your prompt
But Flow uses Veo 3 and supports elements and uploaded frames
it explains why I was able to get 1 with someone just sighing... which apparently is a SFX lol
If you pick a feature hoping it’ll use the Veo 3 model, like first & last frame, extend, or Ingredients-to-Video, it’ll automatically switch to a compatible model instead (veo 2)
this"emotional cinematic scene: sitting around coffee table, Grandpa talking with his 10 years old granddaughter about the meaning of true love Sweetheart, granddaughter ask "what's a true love, grandp?" Grandpa answers "sweetheart, a true love means caring about someone so much that you always want them to be happy and safe, just like how Grandpa always loves and looks out for you." goldenhour, warm soft lighting, beautiful highlight and shadow constrasting."
Yeah, I was trying to get dialogue but there was just loud squeaking at the end 😆
or this one "A tight medium close-up shot focuses on the faces of the young couple, highlighting their emotional exchange. The subject is the woman, her eyes glistening with unshed tears, and the man, looking at her with a reassuring but pained expression. The action is intimate: the man leans in slightly. He says, in a clear, gentle tone, 'I'll write. Every single day.' His breath mists faintly in the cool air. The woman looks up at him, a faint, tearful smile appearing as she replies, 'You promise?' He nods, his expression earnest, 'Promise.' A single tear might trace a path down her cheek as he finishes speaking. The setting/environment is the train platform, slightly out of focus in the background, with a glimpse of a train carriage window. Sounds of the station (distant train whistles, muffled announcements) are present but slightly muted to emphasize the dialogue. The camera is a tight medium close-up, relatively static but with an almost imperceptible handheld feel to enhance intimacy and realism, focusing on their micro-expressions. The lighting is soft, diffused daylight from the overcast sky, skillfully catching the emotion in their eyes and the glint of tears; a subtle rim light from a platform lamp highlights their profiles. The style/mood/aesthetic is cinematic, deeply emotional, poignant, and photorealistic, emphasizing raw feeling. Visual details include the detailed texture of their 1930s attire (e.g., the weave of his tweed jacket, the fabric of her hat), the subtle mist of their breath, and the shine of tears. "
until we can get SFX & Speech in frames to video we'll have to rely on third party like kling 😦
I think my work around in flow is adding more dialogue in 1 clip and cutting out of it in the middle for multi scenes
Anyone found a consistent way to get speech without captions?
Text to video on Highest quality worked for me
only started showing captions when I did frames to video
can you try this :
"Emotional cinematic scene: Grandpa and his 10-year-old granddaughter are sitting around a coffee table. Golden hour light streams in, creating warm soft lighting with beautiful highlights and contrasting shadows. Granddaughter: "Grandpa, what's true love?" Grandpa: "Sweetheart, true love means caring about someone so much that you always want them to be happy and safe, just like how Grandpa always loves and looks out for you." Audio: Granddaughter's sweet, inquisitive voice; Grandpa's warm, gentle, and loving voice. Soft ambient room sounds
but this is the issue at hand @sudden sigil
A tight medium close-up on a young couple's emotional exchange on a 1930s train platform, slightly out of focus in the background with a glimpse of a train carriage window; the woman's eyes glisten with unshed tears, the man looks at her reassuringly but pained. He leans in, his breath misting faintly in the cool air, their detailed 1930s attire (his tweed jacket, her hat) visible under soft, diffused daylight from an overcast sky, with a subtle rim light from a platform lamp. Man: "I'll write. Every single day." Woman: "You promise?" Man: "Promise." Audio: Man's clear, gentle, slightly melancholic voice; Woman's soft, tearful voice with a hint of hope; subtle sound of breath misting; and slightly muted background sounds of a 1930s train station (distant train whistles, muffled period-appropriate announcements, clanging metal, faint train rumble)
THX, I try it right now!
please make sure you chose Highest Quality in the quality option
AND make sure do 1 output for now
😬 i'd be devastated if you chose to generate 4 video per output
Have u made anything animated in flow?
animated as if image to video? or
No like drawn frames
yeah i did
Did you send it?
The animated flow thing u made
still no sound
Cause I wanna see it ? lol
Created feedback for this if you want to comment: https://discord.com/channels/1172568727942860810/1374970038909407282
Also idk if u sent this replying to me but this isn’t what I meant
here it is
cool
@humble zephyr can u screenshot like this?
I’m confused 😭
just wanna see if the tag shows as "highest quality" or not
Is veo 3 better at generation also or does it just add audio do you know?
here
oof , alright i think this is clear enough for the staff/labs team to understand the issues
This is exactly what happened to me. Probably out of the 20 I generated 4-5 did this.
generated over 20 videos, only 4 of them with sound
are we lucky boys?😭
Even when played in the website?
Or just when savrd
website and saved
I keep switching between deep research tasks and flow while I wait for deep research to finish.
Almost out of credits already haha
@humble zephyr have u tried this
actually nvm i dont wanna be the reason why u ran out of credit so fast ahaha
try now
Hi there. I am just starting my Journey with veo3 I am an animator and taking a rather old project and trying to rehash it using veo3. I came across some stumbling blocks with Gemini's understanding of Flow. I ran 6x txt to vids and only managed to get one with audio. The audio part of the prompt looks like this (this was the only thing that worked):
- The girl speaks. Her dialogue is: "And... start!" (spoken in a clear voice with a slight Brooklyn accent). Ensure this dialogue is audible speech from the character and not on-screen text.
- Second issue I guess Ingredients to video does not work with Veo3 yet?
I want the audio and lip-synch and more refined actions of Veo3 however I can't keep consistency unless I can use ingredients. I guess this is a wait and see thing...
Oh and last thing generated ingredients seem to disappear with no way to get versions back
setting
did the video came out with audio tho?
yes! this one is perfect! but why?
and, the picture quality of this video segment is significantly better than that of the previous one without sound. I suspect whether the previous soundless videos was automatically switched to the veo2 model?
I saw a video of someone generate Fortnite gameplay which I thought was interesting
I think the team will clear things up eventually 😄
So for now, just make sure you stick to the prompt structures I gave you for Veo with audio 🙂
I'm trying to extend a scene for more than 10 seconds and using extend feature after adding the generated video to the scene builder. But when I extend it it doesn't generate the dialogues
If I generate these videos separately it, Veo generates new characters every time. How others are trying it?
Wait, I just realized the guy said 'promise' too but in a woman’s voice, lmao
and i think the "ill write " dialogue part were meant for the guy ahaha
haha...yes
either way its perfect
thank you for help! ❤️
How to use it ? It's not available in my country yet
@silver moon
https://deepmind.google/models/veo/ check this out
Not available in India currently so can't use it.
how we can use if change the country or use any vpn ?
If they aren't launching in India, they are missing the most important market of story tellers. Specifically directors like you add a lot of value to the product itself.
Can Veo3’s text-based video generation now effectively maintain character consistency?
I'm trying to use Flow (PRO not Ultra). Why am I not able to use an image as a reference?
Veo doesn’t allow users to upload their own images when they first publish; they can only do so after some time.

So I pay for this, but can do more in AI Studio than what Flow offers...I can't even choose aspect ratio?
😂 Story tellers I hate Indian movies they don’t have physics 😂 😭
I bought Google AI Pro, but I managed to generate ONE video and now I can't create any more. And I can't select 1080p resolution which should be ("included"). Or rather I can't select any Quality, it just say "Upgrade". What am I doing wrong?
Says I've got zero AI Credits
it generates at 720p and there's an option to "download as" 1080p then it does the upscale. It's not as good as topaz starlight though.
idk if 1080p thing is in pro or ultimate only.
Says here it's for both:
https://support.google.com/googleone/answer/16287445?visit_id=638835023451253762-3365861909&p=g1_ai_credit_menu&rd=1#g1_ai_credit_menu
Members of the Google AI Pro and Google AI Ultra plans now have AI Credits that can be used across Flow and Whisk. Learn more about Flow.
veo doesnt really understand negative prompts so if you add the thing you dont want it to do to the prompt it'll probably do it, like if you prompt it to say dont put any bunnies in the background it just sees bunnies in the prompt
Whenever you are in Flow, does it always remind you to subscribe to Google AI?
no
I get this, and when I press Subscribe, it says I already have...
not having that issue, sorry!
Okay, good to know. Seems my account is messed up
I mean, I'm trying to create a video with Veo 3 but it always gives me an error, even though I have credit, is anyone else having the same problem?
I will have to try later again, can't make any generations and my account is not detected as Pro...
Consider reporting it in #1319424187730755636
Thanks, will do. Actually saw multiple people having same issue.
I have the same issue 😦
are you from US
I saw multiple people from US having this issue as well (in the bugs channel). But no, I'm not from US, I'm behind US VPN.
im not from US either, but i can access flow
Using VPN or not?
yea
i mean i only connect the vpn just for short time , then i disconnect em , i can still generate stuff
Nice, lets hope they sort this out soon.
Can you choose to generate only 1080p videos?
I read that they are included and not consume credits, can you confirm that?
Members of the Google AI Pro and Google AI Ultra plans now have AI Credits that can be used across Flow and Whisk. Learn more about Flow.
no thats not how it works, you can only upscale a video to 1080p
Ohh, so you generate 720p (which then freely gets upscaled), which will consume how many credits?
#flow-general・🌊 Warm, cinematic voiceover tone]
“At the heart of every home… there’s a kitchen.
And at the heart of every kitchen – there’s Liam.
For over 12 years, Liam Kitchens has been crafting dream kitchens in the city of Tiberias –
where Italian design meets precise carpentry, where every drawer glides in silence,
and every finish speaks elegance.
This isn’t just a kitchen.
It’s where your mornings begin, your dinners are shared, and your memories are made.
From custom-made countertops to lighting that warms the soul –
we design with heart, build with passion, and deliver perfection.
Liam Kitchens – the heart of your home.
Schedule your design consultation today – and bring your dream kitchen to life.”
hm i forgot how much
no it doesnt cost any credit for that
same issue
Same issue here too 😦
how are people creating longer videos with consistent characters with dialogue, I can only get 8 seconds and in scene builder it doesnt let me use veo 3
please help 🙂
Is there a sample video you mentioned that shows how they managed to keep a consistent character using Veo 3?
Veo 3 can create AI videos with multiple characters and voices.
X: https://x.com/jerrod_lew
Instagram: https://www.instagram.com/jerrod.lew/
LinkedIn: https://www.linkedin.com/in/jerrod-lew/
i guess its just 8 second clips cut together lol
Yeah obviously
but am I right in assuming you cannot extend VEO 3 in scenebuilder?
Yea u can’t
Thanks for all the Flow love! We're still in the early days (how is it one day since launch
), so we appreciate all the feedback! While we work on fixes, wanted to clarify two things:
**(1) Flow is currently US-only. Other countries coming soon! **
While Google AI Pro is available globally, Flow is currently available to US based users only. You can still enter the tool (with the grey bar at the top telling you to subscribe) and look around, but if you try to generate you'll get a generic error. We're working hard on expanding to more countries, stay tuned!
**(2) Veo 3 is Text-to-Video and Frames-to-Video (first frame) only. If you select a different feature (first & last frame, extend, Ingredients-to-Video), we'll switch to a compatible model and ask you to confirm **
Veo 3 is still early and doesn't have all the same features as Veo 2. To help you complete your action, we switch you to a compatible model and ask you to confirm the switch. We'll make this clearer in the future, but for now check your settings panel before you submit and see our FAQ for a table of model capabilities: https://labs.google/fx/tools/flow/faq
Thanks for all the feedback and being such a great community!
Google Lab's AI video generation tool, powered by Google DeepMind’s latest model: Veo 2
ok cool
but first frame works?
it was giving me problems, maybe its because i already got the message once and i had to restart the project
From faq page
Veo 3 is currently only available for Ultra subscribers, and supports:
Text to Video, with speech and environmental sound.
First Frame to Video, with environmental sound.
Note that Veo 3 is not yet compatible with first + last frame, camera controls, extend or Ingredients to Video capabilities. We will default you back to Veo 2 models when you want to use these features.
Anyone know when image upload will be available for flow?
Is Korea excluded from flows available country list?
@sharp dome ^^
how can I make my veo3 talk? I cannot get any meaningful words to come out of my AI
White background. Camera focused on mans face. man says " Hello world I am AI"
man says "Hello world I am AI"
man says hello world I am AI.
etcetc
I cannot get it to work
Try this prompt: "A clean white background with a cinematic focus on a man's face. The man has a confident expression, looking directly into the camera. He is speaking clearly, saying: "Hello world, I am AI." The lighting is bright and even, giving a futuristic, professional feel. Close-up shot, high resolution, realistic skin texture."
@hardy halo hey swedish output, and language veo 2 is very bad and eat up my videos promt is there anyway to fix this or beta or something can someone look it to this? On Gemini advance pro edition
Swedish output? What’s that mean
If i buy Google AI pro, will i have the full optimal Flow, or am i not getting VEO 3 for example? also it should work well if i buy it and im on an VPN aswell right?
A clean white background with a cinematic focus on a man's face. The man has a confident expression, looking directly into the camera. He is speaking clearly, saying: "Hello world, I am AI." The lighting is bright and even, giving a futuristic, professional feel. Close-up shot, high resolution, realistic skin texture.
Iam almost out of credits
and haven't got a single cohesive word
with all the videos I got, I can make video titled
"Veo3 Speaks! (gibberish)"
Prompt issues that’s all I can say
genrate a random video
they advertised that they released an ai model video generator that can speak but its not delivering. and its supposed to be my fault?
It is
In a way ,
If someone can give me a prompt that can actually make the AI say something I will not be annoyed
I refined your prompt
"A cinematic close-up focuses on a man's face against a clean white background, his expression confident as he looks directly into the camera with realistic skin texture, lit by bright, even lighting for a futuristic, professional feel. He speaks clearly, saying: "Hello world, I am AI." Audio: Man's clear, confident, and articulate voice, with a neutral, slightly synthesized or polished quality to subtly hint at his AI nature, against a backdrop of near silence or a very faint, almost imperceptible futuristic hum."
Good luck
Starting to get worried that people don’t take the whole trial-and-error process seriously, they just expect everything to be perfect on the first try 🫠
Anyway @sharp dome let me know if the prompt I gave ya working for you or not
there is a fine distinction between prompt skill issue and an underbaked incompetent AI.
I am versed fluent in sora Midjourney and that chinese stuff and other text based prompts.
I get it that u r trying to defend this veo3
but unless u urself can actually give me a prompt to make veo3 speak, something I am sorry but I can only say the google over advertised and their AI is not ready
I just did
Please set your anger aside for a moment and actually read the chat properly.
r u sure? u gave a prompt that made it to actually speak something not gibberish? let me give u back ur result
Is this the new one I gave?
yup
How can I get help with flow, I got the Google Ai plan and still cant get access, it keeps on mentioning to subscribe, I already did.
(i am starting you with dead expression while ur veo3 speaks in gibberish for 10 hours no matter what style of prompt it gets)
can u run it on ur veo3 and give me back result? I am honestly tired of hearing gibberish
this is sooo disappointing
I don’t have ultra plan
Try this
Close-up shot, futuristic professional setting: A man with a confident expression looks directly into the camera against a clean white background with bright, even lighting. Realistic skin texture, high resolution. Man: "Hello world, I am AI”
u can run veo3 with pro?
