#flow-general・🌊
1 messages · Page 13 of 1
so if your upload was denied because of restrictions such as giving an image to vio 3 fast or using ingredients with veo 3, it will kick it back into veo 2
which i dont like
How woudl I know it is deinied if it genreates something?
ive noticed if you do AI meme videos of a niche that no one else is taking advantage of, its likely to blow up on instagram 👍
So when I do it I click Veo 3 but output it gives me veo 2?
good for gaining real growth without bots
can i see the prompt
that u were trying to do
Ya sadly says Veo 2 even though I select Veo 3
Sure one sec
and was it text to video, or frame to video?
Text to video. This was the prompt: The camera is placed at the bottom center of a crib, angled directly upward, fully surrounded by vertical wooden slats on all sides. The sound of gentle baby gurgles and coos is audible as a soft. Soft daylight spills in from a diffused window on the side filling the scene with a hazy, overexposed glow. A silhouetted woman leans into the view from the top of the frame, her body arching over the crib, completely backlit so her facial features remain hidden. We don't see her but her hand enters slowly from above, fingers open and gently waving, hovering just out of focus. Babbled voice says the words Ma...ma...
Trying to be the POV of a baby in a crib basically
oh
.. wait let me try
i do know that.. veo 3 wont generate audio that involves children
"baby" maybe
yeah
Ahhh maybe that is it
say new born, wrapped up
but it can still generate with veo 3 just no audio
still
Ill try it but anything else you would change in my prompt
ill test for you if youd like
Sure thanks! Let me know what you do as well to ge a result
ye i can also try to refine it
Appreciate it
do small changes to figure out what it is
The problem is, I keep burning credits testing when I think I have it setup
yeah lol
is tehre a place we can TEST prompts before running?
I just had that pop up for me wasnt there yesterday
I guess I would love to see both of your results and I will see how to proceed
yep veo 3 fast, it generated but no audio
From a worm's eye view camera angle positioned at the bottom center of a baby crib, looking directly upward. The frame is entirely enclosed by sharp vertical wooden crib slats, creating a cage-like perspective looking out and up. Soft, diffused daylight spills in from a side window, bathing the scene in an ethereal, overexposed, hazy glow, heavily backlighting the figures from above. A silhouetted woman leans into the top of the frame, her form arched gracefully over the crib edge, her features obscured by the strong backlighting. Her hand slowly enters the view from above, fingers gently open and waving just out of sharp focus. Soft, gentle baby gurgles and coos are audible, accompanied by a soft, babbled voice whispering the words 'Ma...ma...'.
I basically get the same thing, and again no audio
Was wht you generated above veo 3 fast or quality? Any idea why no audio? Word baby?
veo 3 fast, the reason would be , its because it involves baby
in all honesty they need to put Flow in an electron app
Any tips on how to get around it? I wish it would tell you Hey this word you are using wont make audio. But even then at least give the cooooing ahiiing of baby
with audio, had to refine the prompt
A worm's eye view shot captures the scene from the bottom center of a crib, looking straight up. Vertical wooden slats surround the frame. Soft, diffused daylight spills in from a side window, creating a heavily backlit, overexposed, hazy glow that fills the space with an ethereal ambiance. A silhouetted woman gently arches into view from the top, her features hidden by the backlighting. Her hand enters slowly from above, fingers open and softly waving just out of focus.
lmaoo
I mean even then the mom is saying it lol
LOL remove the baby
God damn baby ruining everything lol
you can tell it that there was a child like goo goo and it should work too
let me try with,, a young male voice saying goo goo
im gonna cry if its actually a man voice
lmao
lol i appreciate you guys using your cerdits on this with me
i literally bought this to mess with it
i dont mind tbh , thanks to veo 3 fast model
You get ultra? I am on pro debating if I want to upgrade
im on pro, i think @analog zealot is on ultra
honestly not sure of the difference, wrigley confirmed something about a pro version of the model tho
I’ve only started with this baby stuff and every generation has been bad so far 😩
guys my country doesnt allow veo 3 and i have a test subscription witch country allows it?
usa lmao
a young male voice whispering "Ma...ma..." did not work
damn
Did it reject it?
gimme some time
What’s it look like?
it looks .. like the mom got deformed arm
Lol saying John Cena you can’t see me gesture
Hmmmm alright I’ll try some more. I think I should do one prompt that 100% will give me a good output so I’m not going crazy
lets see if a Lyrebird works
just make sure you chose the right model
I 100% picked veo 3, after the generation it says veo 2.
Yup, report where here or on the site? Does anyone here even read these?!
well bugs are being flooded in the #1319424187730755636 but im sure yours is a concern
wonder if they reimburse
Lets see! let me know the prompt too
That one is not bad other than the crib being deformed haha, let me know the prompt.
Camera Placement & Composition:
The camera is placed at the bottom center of a baby crib, aimed directly upward. The view is completely enclosed by vertical wooden crib slats on all sides, forming a symmetrical cage-like frame. The shot is static.
Lighting & Mood:
Soft daylight filters in from a diffused window off to the side, creating a hazy, overexposed glow across the scene. The light washes out detail, lending a dreamlike or unnatural calmness.
Audio Design:
In the distance, there's a quiet ambient noise of a TV playing, muffled and mostly unintelligible, suggesting another room.
Footsteps begin softly—slow, deliberate, like someone walking barefoot or in socks on a hardwood floor—getting gradually louder as they approach the crib.
Just as they stop, a gentle creak is heard from the floorboards, timed with the presence entering the frame.
Action:
From the top of the frame, a silhouetted woman leans over the crib. Her body arcs gracefully inward, completely backlit—her face obscured in shadow, unrecognizable, only her outline glowing from the soft daylight.
Her hand enters the frame slowly, fingers open and gently waving, just slightly out of focus due to shallow depth of field. The motion is deliberate and calm—maternal, but uncanny.
Sound Cue:
At that moment, from an unseen Lyrebird, we hear a soft, babbling mimic:
“Ma...ma...”
The tone is slightly off—too close to human, but not quite.
ai to write an overly detailed prompt works good
What AI did you use? Sometimes I try with chatgpt it is lost
got one you recommend?
i used chatgpt 4o
what was the prompt to get it to output or do you have a GPT of it?
i always try and start off with "make a prompt for veo 3 that describes...." or something similar
GOt ya
Getting there, would like the ma ma said more towards the end but I think I am just going to edit audio myself
How are people extending there videos over 8 seconds whilst keeping the same scene.. I’ve seen loads on TikTok but I’m struggling
scenebuilder > extend
You know what I mean.. like people are making yeti videos and I don’t get how there doing them
Wasted like 4k credits messing about but I can’t get it right
so i think what theyre doing and what i thought about doing was generating with veo3 getting screenshot of last frame and using the frames to video feature with veo 3
Thank you I’ll give it ago now
otherwise youd have to use a template for each clip that describes in great detail the characters
which i feel like would be more difficult i havent tried
use the starting frame
It come up with this message when using the frames to video
ignore the rest
Then it takes you back to vo2
set to veo3 quality
frame to video, set first frame to the last image of the previous clip and dont fill in the rest
it has to be quality not fast sorry
Ahhhh I’ve got it I think
then put a prompt describing how youd like to continue
tbh idk why this isnt already implemented lmao
You a legend I think I’ve got it
lmk how it turns out, ive never actually tried
I’ll let you know brother!
Does f2v work w audio now
yeah with quality only, not fast
When do credits reset for pro and ultra? If you use all credits can you buy more?
and u cant have an ending clip, only a starting clip
buy more or wait till ur subscription renews
i somehow got audio in fast img2video veo 3
Link for veo 3
Google Lab's AI video generation tool, powered by Google DeepMind’s latest model: Veo 2
WHATTTT
nice, everything's developing very quickly.... they probably earned a lot of money
so thats what was setting dudes thing back to veo 2
Thanks 😊
@crisp basalt if youre in frame 2 video with everything filled in and switch it to veo3 fast after you set everything else it will generate but in veo 2
bug
show the setup of your prompt and images
i think it was 20 credit
yeah so its listed wrong on mine
its using veo 2 for some reason
wrigley is trolling, he has early access or sum
@crisp basalt is the date of the commit 6/4/2025 in the user drop down
the wha
this?
....
why
idk it wont let me do what youre able to do
i thought u can
i can only do it with veo 3 quality, not fast
Explain to me what is happening lol
were u using frame 2 video when it was doing it
Always text to video
i guess ill be taking a break for a little bit till they get it sorted with the model selection then
and yet still no extend or jump to Veo3... I know there's workaround etc, but it would be nice to have an almost complete one tool for making longer scenes with audio
veo 3 can do jump to
the audio transferring?
yeah environmental audio
it cant do speech because jump to uses image to video as workflow
they stated it in the faq that veo 3 imgtovideo first frame cannot do speech audio at all
used jump to feature
nice, thanks for that info, I think last time I was testing it ,is just two people talking, so that's why I didn't get an audio
no prob , i do hope they bring the speech audio too for img2vid
because their deepmind page literally got vid example of veo 3 extend scene video involving two character talking
new toy to play, new money to waste, lol... but at least cheaper than quality
it becomes enjoyable for me to use flow now cuz of the cheap veo 3 price for fast mode ahaha
But are the outputs good quality?
Is there any way to realistically render animal hands/paws?🤣
How do people get a frame to go over 8 seconds by adding them together (extending it doesn’t work) I’ve tried to do frames but you can only do “quality” and it doesn’t do the diolog voice over
Is this a case of describing the character from head to toe? As well as the environment?
wait image to video has audio now?
Hello guys, it's possible create a vertical videos in flow?!
please,add more frames and ability to change duration of animation (between them). cause when you need just a simple transitions between two frames often result in 8-second, somewhat chaotic animation between the initial and final frames, which is often too long for quick, smooth sequences.
also please make a prompt field resizeable🙏
So veo 3 turbo is here 🔥💯
It's seems so
veo3 fast is good,but why cant we extend this animation? its not fair.it means that you can only generate short 8 sec clips(
maybe it can be used just for a preview/ what the difference between fast and quality veo3?
You cannot extend with fast?
no. only with veo2 fast..but the quality of animation i way better with veo3 model
you can't extend with veo3 in general
Oh yeah, I hope this is the next feature that come
We need the extend and jump to for veo 3
when they release extend for veo3 it's gg. 1 hour long movies will be viable
i think 90% of all flow users use it for prompting text2video
thats why generations all pretty much similar
pure dataset
Un robot solitario che guarda il tramonto su Marte, atmosfera malinconica
Guys, I'm trying to generate a video, image for video, using veo3 - quality, and there's no sound coming out.
This prompt got sound working for me, only in VEO3 though - The audio features the rhythmic sounds of the stream, occasional splashes from the fish.
I'm trying to generate a woman speaking, but it's silent
So for frame to video for veo 3 quality, it seems it will NOT allow talking
Any prompt that has things asking for sounds will work “ the sound of the valley was humming “
Any prompt that is asking for speech or talking will not work.
Either this is intentional to avoid deepfakes or it’s a major bug. Either way needs some clarification from Google.
@main saddle
It’s just not available with Veo 3 yet same for ingredients
It is in Veo 2
Yes it is
I’ve been using all of yesterday
You can’t have dialogue with any of them
With sound and all
Yes no talking but veo 3 quality is selectable and with sound
I've deff seen VEO3 with talking
Yeah well in my case I need dialogues
Yes there’s dialogues in Veo 3 lol just not when using ingredients/ frame to video / extend
Hi all. New here so sorry for just dropping a question straight away but I've set out to make a music video for a song i produced. The issue is, most of the images of my generated female singer have been rejected and I can't understand why. They're not even slightly controversial IMO.
Yes thats always been the case but previously Veo 3 was not able to be used with frames to video
It now works, just sound though, no dialogue.
So either this is a bug or its intentional to prevent deepfakes e.g uploading an image of elon
e.g this is a frames to video from the image here.
15 generations and it will not give me a dialogue at all, happy for sounds and music though.
how do i remove the watermark?
Signed up for the Ultimate subscription but it seems I'm stuck at the loading screen. Has anyone had this issue?
im pretty sure the videos yall are seeing online are not more than 8 second clips, theyre cut and they do not look the same
well i cant do veo3 fast frame to video still so idk how you managed even environment sounds
Of course
I mean they don’t need to clarify it tho, it’s already stated it in faq, it is up to user if they want to check the faq or not
But , I agree, there should be some kind of indicator or tooltip that warns users if a certain action wont give the expected result. Like with Veo 3’s frame-to-video, users should know it won’t include speech audio, just environmental sounds
That “switched to compatible model” toast , not the most helpful in my opinion . I’ve seen some of peeps in here going full detective mode (or rage mode 😇) trying to figure out why a feature suddenly disappears. Would be nice if the message gave a quick reason or heads-up, just to avoid another round of “what the hell just happened??” in this channel
Why would it be obvious to a user that veo 3 doesn’t include speech for frame to video
I mean, is it really that obvious? Or did you just skip the FAQ page and decide to roll in here with wild theories instead 🤔
I think veo 3 fast frame to video was a bug that I got last night lol
I can’t do it anymore now
...
A good bug
So u got bugged access
Well I think ur outcome was the intended tbh, why would it still generate a veo2 when Veo 3 is selected you know
Are you dumb
You don’t need to glaze Google
Unless it charged u 100 credits then it was just a quality frame 2 video instead of fast
It’s poor ux
Wasn’t glazing but calling me dumb won’t make u a smart person either lol
🤷♀️
Hi guys, I have a problem and I don't know if you could help me.
I'm generating scenes in Scenebuilder with Flow, and the scenes are created but they don't have audio. I don't know how to make them appear with audio. I've already adjusted the prompt, but the only scene that has audio is the first one. The rest of the scenes generated in Scenebuilder don't have audio. Does anyone have any advice?
Says one is selected when another is
The feedback channel exists for a reason. No need to unleash your anger confusion here just because you didn’t feel like checking the FAQ first 😅 Calling someone dumb over that? C’mon now.
Apparently you can only do environmental audio and not voices in extended clips
You’re glazing. I’m entitled to comment in the general channel
“Jump to” only
What does jump 2 do
It’s self explanatory in faq page
The video of the motorcycle wasn't jump 2 tho right?
Basically create a new scene that maintains consistency of character/scene something like that( which in my experience it’s a hit and miss most of times)
It was jump to
Like I said it’s a hit and miss,
So what good bug did u experience if it's normal
By not overly micro analyzing everything
Oh well, still bugged regardless
And did you report the bug yet lol
Good!
Wish you would be more help with how much u talk
Yeah just ignore him, he thinks he’s a mod
Yep buddy you contribute nothing to the overall discourse
It's really just how I asked him earlier and made it extremely clear that I was using frame 2 video and he never once told me I need to use jump 2 😂
gotcha
Just heard Veo 3 now has a fast option for 20 credits, 😮
it is whisky not working? I was using all day and now it just stopped
Hi i just subscribe to Google AI Pro and i keep getting this message on Flow "Flow is available exclusively to Google One AI subscribers. Not a subscriber yet? Take a look around." Any ideea why and what can i do? Thank s
"Create an intro video with the name SSF in the style of the Universal Studios intro."
Message #flow-general・🌊 create an intro video with the name SSF using UNIVERSAL STUDIOS logos template
Hello all just joined, wanted to ask for some insight towards how to create a character and keep the character consistent throughout ?
its literally just veo 2 with audio defeats the whole purpose of veo 3
@dusty dust I'm struggling with this litterally right this second
have a super detailed description for your character work's for me
how does your prompt look like? i literally just had gemini just create a prompt for me, but i see in another section #flow-tips-n-tricks・✏️ it looks more like screen play
Big part of Veo3 experience was the audio, so if I can create more videos with audio for less I take that as a win.
@dusty dust this is how i use it in a practical sence and this works fine
A pale young woman with shoulder-length black hair, dressed in a loose hospital gown with electrodes and medical wires attached to her scalp. Her skin is cold and bluish from extended unconsciousness. Her expression is always blank or subtly off, and her movements are slow and unnatural, as if controlled by something else.
keep in mind im using veo 2
this is a prompt im using for an ai short film im working on called DONT WAKE HER
here a different prompt that gave me this result
PROMPT: A pale young woman with tangled black hair and electrodes attached to her scalp lies unconscious in a hospital bed. Suddenly, her eyes snap open — fully black with no visible whites. She remains completely still, her expression blank and unsettling. Her skin is bluish from long unconsciousness. Harsh clinical lighting flickers overhead. Cinematic horror realism, 4K.
is there a way to establish a prompt to a name? and just bringing up the name? in #flow-tips-n-tricks・✏️ i see another user , writing completly different
how does it know who "Yogi" is
it kinda work actually
o
damn this looks so real
and that was veo 2 😄
wut!
veo2 fast?
it was fast yeah
holy?
thought it's veo 3 xD
When using Veo 3, Flow arbitrarily generates videos in 16:9 or 21:9. Adding the aspect ratio to prompts has no effect. Is there any way to lock in the aspect ratio?
Did you try telling it the requirments before the actual prompt?
If I cancel my Ultra plan after a month and resubscribe later, will I still get the discount?
very unlikely. Veo 3 is definitely beta quality atm.
hello i want to asking. how to get credit in veo 3? when im clicking button subscribe it reddirecting to page
You're already subscribed
You can access your membership benefits at any time. Learn how to manage your Google One plan.
but my credit still 0
Hello, i need a little help for generating clips in veo3 for a Video - My prompts not good enogh - Can someone help please asap
Timon and Pumbaa went looking for the legend of Elsa living in the jungle — and let's just say, they weren't ready.
Was it real? Or just a frozen fever dream? ❄️🌴
Either way... Elsa's got WiFi in the jungle now.
Don't try this at home. Or do. We’re not your mom. 💀
#AIgenerated #veo3 #JungleElsa #TimonAndPumbaa #FrozenInTheAmazon ...
@quasi gate This video helped me get started: https://www.youtube.com/watch?v=oWSprcjRJq8
#AI #aivideo #Veo3
Think AI video can’t look cinematic? Think again.
In this full tutorial, I’ll show you how I used Google's new Veo 3 to create movie-level videos with just one prompt — no editing, no VoiceOver.
The recent popular street interviews, talk shows, auto show interviews, etc. are all AI videos generated by Google Veo 3.
Y...
i send you a dm
فيديو عن سفينة ضخمة فالمحيط وطقص عاصف رياح قوية وامطار غزيرة
can you send the prompt?
the sky and day and night cycle accelerates in speed. people in the surrounding area are confused and scared. the sun has sped up so much that it appears as a thick glowing line in the skying while the clouds move fast. the buildings worsen fast because of how fast time is moving
hello guys I'am a beginner
how to creat in video AI
#flow-tips-n-tricks・✏️ A cinematic high-stakes police chase after a bank robbery in a modern city at night. The thief, wearing a black mask and leather jacket, escapes in a sleek black car under heavy rain. Flashing red and blue lights reflect on wet asphalt as police vehicles pursue at high speed through narrow streets. Dramatic slow-motion shots, aerial drone views, and intense close-ups of the thief’s determined eyes. Moody lighting, intense soundtrack, gritty urban environment, rain splashing on car windshields. Directed in the style of a Hollywood action film.
hi guys, ive just arrived here still learning , theres lots everyday upgrading
How can I generating videos?
Google Lab's AI video generation tool, powered by Google DeepMind’s latest model: Veo 2
Are the subtitles still appearing or have they been resolved? Not generated anything for a while
Has anyone else’s credit consumption changed for fast veo 3? It says 20 and it’s not doing sound anymore.. also quality is 100 but there’s no sound
I have only 10 credits now xD
#flow-general・🌊 message This style of prompting worked for me for 10 or so Fast gens with correct speech and no subtitles.
Great so leave off ‘he says’ and quotation marks and word it as though the character is expressing something and it will work? I’ll give it a go later and report back, thanks
Hello, why it tells me that I have to be a Google AI subscriber when I already am a subscriber?
where is flow available, which countries ?
because it says it is not availiable in my country yet
Current available countries is in the FAQ. https://labs.google/fx/tools/flow/faq
Google Lab's AI video generation tool, powered by Google DeepMind’s latest model: Veo 2
its not available in UAE yet
who can i dm for some help...
@crisp basalt did you use the quality version of veo 3 when you did the frame to video or "jump to"
thats literally the only one that will work for me with frames
;( i dont wanna spend 100 credits for frame to video with environmental audio lol
is it possible to make veo 3 output other language?
what prompt to use? because mine is saying other language is not supported? how can i make the character says other language?
@wanton mulch do you know anything about veo 3 - fast and fame to video
yes, my issue is that i can only generate frame to video with veo 3 - quality
it says frame to video + veo 3 - fast isnt compatible
this is someone using fast with a provided image, its impossible for me to do the same
@wanton mulch
what may be???
i either should be able to use veo 3 - fast + frame to video or not. like someone can test and just give an answer maybe
i can use veo 3 - quality + frame to video but thats 100 credits
Having a hard time it may be scene specific with veo3fast directing a first person video it wanted to make it a third person it is great for some stuff with zero flaws though. It's also possible I would have the same problems with the main branch as I am trying to see if I can get by with my content creation by 5xing my gens
someone literally sent a screenshot of it working with the exact same image that im using
ive used a prompt that doesnt involve people and it gives the same result
lol i just found out theres a veo 2 version of flow for free https://docs.google.com/videos
Access Google Drive with a Google account (for personal use) or Google Workspace account (for business use).
شكرًا للتوضيح! بما أنني لا أملك الفيديو فعليًا، سأقدّم لك وصفًا عامًا ومفصلًا يمكن تعديله لاحقًا إذا زودتني بالمزيد من التفاصيل:
وصف فيديو: سفينة في المحيط
يُظهر الفيديو مشهدًا رائعًا لسفينة ضخمة تُبحر في قلب المحيط. تحيط بها المياه الزرقاء العميقة من كل جانب، والأمواج تتلاطم بلطف حول هيكلها. السماء صافية أو ملبدة جزئيًا بالغيوم، مما يضفي طابعًا دراميًا على المشهد. السفينة قد تكون شاحنة حاويات، أو سفينة ركّاب، أو حتى سفينة استكشافية، ويتضح من حركتها الثابتة عبر الأمواج مدى قوتها وثباتها وسط الطبيعة الواسعة. ربما يُسمع صوت هدير المحرك أو صوت الأمواج، مما يُضفي لمسة واقعية وهادئة على أجواء الفيديو.
إذا كان لديك تفاصيل إضافية مثل نوع السفينة، الجو (عاصف، غائم، مشمس)، أو ما إذا كان التصوير بطائرة درون، أخبرني وسأخصص الوصف أكثر.
#flow-general・🌊
how can you generate a video there?
go to it, make a new video project and pick the first or last template, on the side you can click to gen a vid
cant find it
where should i pick a template when created a new video
hmm...i cant see that. maybe i need a vpn
nope.aint working for me( so you can generate on a timeline like a flows scene editor?
without limit? gemini is free also, but you have a limited video generation per day
no limit so far
oh nice
halp I just ran a prompt, it got to 99%, then I refreshed the page and the video disappeared
pretty sure i actually paid credits for that
How do you have speech?
idk, speech is external tts i think not veo 3
So Veo2, so no audio right?
yep
Still amazing if no limit, good way to try your video first there before using it with Flow/Veo3
Sometimes, if you keep refreshing , it will appear, sometimes it will appear in a much lower lines of your generations
vpn my guy
Hi, im just new using flow. i encounter flow video has no music or sounds. is this normal?
whatre u trying to make
it says i need the ai pro but then i subsricbed and after it still doesn't work can anyone help
damn, thank you, I feel like I could have save a lot of money/AI credits, if I have known that before... now I am wondering, what's the catch? lol
just tried to create a basic video with script and scenes but it generated without sounds. is it normal?
yes if youre asking it to do something that it wont allow you to generate
flow only for video without sounds?
u can reword it and use different key words to get similar results to what u want but with audio
are you using veo 3 or veo 2
how i can know if veo 3 or veo 2?
you select the option
i just go here: https://labs.google/flow/about
oh wait let me check
oh im using veo 2
let me re generate might be due to this
wow nice its the settings! thank you for the help! @analog zealot
same
dialgoue with more than two people gets messed up pretty often, no wonder everyone's doing street interviews
is there a sulotion
havent found one yet
if you are in trial version beware that in all generating attempts it will use monthly points
Hey has anyone figured out how to use the reference image in frame to video or ingredients to get speech? I need to use the same character
you can't
So what’s the point of flow if you you can’t “flow” 🤦🏽♀️😭😅
i wonder if they could add a section for characters that were allowed to generate.
and how they sound and personalities
me about to burn 8000 credits on a 10 minute video 🥹
tf is it about lol
you can still maintain consistent charcters with prompts
but yeah they probably want to avoid the deepfake problem for now
original tv show episode
but multi-character dialgoue is cooked beyond 2 people so it's prob gonna cost even more just to get some acceptable results
omfg
What are you saying in the prompt. I have it describing the same look but a different person comes out. I’m not uploading a pic just describing it
I want to use the same character it gave me
if they had a dedicated way of creating characters then that wouldnt be a problem
give ya something
in my case i'm using really distinct looking characters with specific outfits. e.g Blue haired japanese man with blue suit and brown eyes. Even though it'll change every scene, the viewer will know it's the same person because they stand out
But in the face they don’t look different? I’m describing the same outit and say black man and it gives me different looking black men in the same outfit lol
gotta define facial features
Yes still different face, shade of black skin
ive noticed like on the yeti videos, they dont completely look the same but theyre close
Ok
today I learned
have you reached the point?
yep, 20 vids a day
nice
woow wsp
Ultra has no limits I think
if it does, i'll find out in a bit lol, going to be doing at least 100 gens
from what I see Ultra only has limits on a matter of credits, and you can purchase more if you use your 12500
for some reason.. when I use an image as reference for veo 3 the audio is gone, is this a known issue?
apparently you should be able to use frame to video with Veo 3 - fast but i can only get it to work with veo 3 - quality, you can only do environmental sounds like a person breathing making a noise but no talking
"SFX" pretty much
ohh.. but are you able to extend with veo 3 quality? cause it is changing the model to veo 2 here .-.
like i said its bugged for me so idk
No matter what I do I cant get sound. Wasted almost 1000 credits with different prompts and still nothing I'm losing my mind.
x2
text to video, veo 3, quality.
is that why lol if anything invovling kids even if they are animals?
use teenager
yeah it randomly doesn't do dialogue sometimes. rip
its not random lol
it makes sense if the prompt says kid or kids that it wouldnt use audio
ive "diagnosed" a prompt audio issues because it said baby
or create audio
yea true there are some cases where it seems to never produce it, but also sometimes when there's no problems it will have no audio
yeah adding environment sounds instead of just specifically voices helps get it going
this is what happens when u do frame to video with veo 3 trying to get it to talk lmao
gotta love when you make 4 variations, one has good audio, another one has no audio and the other 2 have the most random audio with no speech at all xD
yeah its a pain
Cinematic shot in a snowy pine forest, late afternoon. Camera pans forward as Betty the Yeti (tall, white fur, pink scrunchie, baggy flannel, yoga pants) walks confidently toward camera holding a steaming thermos. Behind her, two light brown furry teenagers (half yeti, half bigfoot) tumble through the snow, shrieking and body-slamming each other into a log. The camera zooms in mockumentary-style as she speaks directly to it. Dialogue (Betty, smiling dryly): "Hi. I’m Betty. I’m a full-blooded yeti... single mom to these two disasters. And yes—those are mine." (In background, one teenager yells: “I FOUND A DEER LEG!”) this is the prompt i still cant get audio
video looks great though
thats the no audio prompt?
yes
bruh what in the world is this nightmare fuel
yes I mean they are big foot so they look like teen big foots lol
are you telling me this is not real?
anything is real
about 3/4 done with the script and now we broke
do you have any tips to keep the voice consistent? for some reason on my second clip with the same instructions for the voice AI started singing, I appreciate the enthusiasm but need more consistency xD
i havent messed with that, i guess explain the emotions
and their age maybe
i think the yeti videos use voice overs tbh
lmao
I'm trying to do a reverse shot scene extension....any suggestions? I'm using scene extender...Prompt "Reverse shot using wide angle lens, still raining. "
Yo is veo just a text to video type thing. Like if i send it a picture would it be able to animate it?
yes, but veo 3 does ai audio
as long as you are compliant with the rules, it will use your image as base, yes
image is only for veo2 still right
appreciate it guys.
with what
i cant help with that my man, maybe paypal will work for converting funds
there are ways
Can I get some feedback for my most recent project? https://www.tiktok.com/t/ZP8Mt2o3Y/
Hi Jeetu. Not sure if there is any restriction on your country, but I'm in Colombia and here VEO 3 can be used, restrictions are mostly for google 1 stuff and youtube premium
getting failed generation for having "girl" in prompt instead of "woman". gg no children gonna be in these movies.
well.. the thing is.. I would have to use a credit card for that, so I don't see this really viable :C
just a question why Flow is generating video of people speak in english only no other languages and most video i generated have no sound ?
no no, I totally get your point, what I'm trying to say is that.. in order to get the sub the account needs to have a credit card linked, so I think you can rather pay some one to generate things with your prompts and send to you or sth
you need to specify but it is creating videos for me speaking in spanish with no issues
sometimes it can go and use another language tho
beta feature things
idid tbh but it start generating at the end it says Error generating always
I don't know if live chat support can help checking the specific error you are getting but that's what I would do.. maybe give it a shot?
could be just wanted to see if iam the only one having that issue
Looks like Veo 3 fast, dont support image to video yet
well.. same is happening to me also in quality mode when I'm using flow
btw idk if u can but can u give me the prempt u generated that video of people speaking in spanish just wanna test
"Abstract, flowing cosmic visuals. Vibrant nebulae in hues of deep magenta and warm gold fold into each other, creating nested, womb-like patterns. Star clusters form infinite fractal expansions. Organic, yet cosmic and infinite. Dreamlike, ethereal. Audio: Male voice-over, questioning and wondrous, with emotional echo saying 'Pero… ¿y si nunca estuvimos separados del universo? ¿Y si estamos contenidos unos dentro de otros… …como un útero infinito que no deja de parirse a sí mismo?' Music swells, becoming mystical; soft, pulsing cosmic heartbeat SFX. aspectRatio: '16:9' durationSeconds: 8"
here is one
appreciate it man
Curious where are the Google sponsored FLOW tutorial/video/resources? Don't say here....
did i spend $200 making a 11 minute tv show episode? yes. am I bankrupt now? yes. was it worth it? yes.
I think I have spent about 600 to 700 expirementing to get 100% identical characters and voice. So far I have nailed the voice. The look and clothing. The geometry of face and facial features. I’m so close
Flow is somehow not working for me but Gemini VEO 3 is doing its job, I just wish we could have unlimited video generation, while you get used to it it runs out 😢
Yeah they better get to work with how much we've paid them
yea but it seems like Veo 3 must consume huge computing power
since it generates high quality visual scenes and audio altogether
for example, it tries to sync the video and audio I think this will cost insane computing power
Definitely, but I mean I have the google pro plan 20$ a month and imagine 10 videos 8 sec each is nothing, the first day I started using it, I couldnt even figure what prompts to use to somewhat get it right
If there's not some revolutionary advancement in Computer science, I think we are going to suffer with this high cost for a long time
Honestly its worth way more but I guess we gotta wait for some chinese company to come out here with some copy for cheaper until we get more videos and longer ones haha
I think they allow Pro users to use Veo3 because Google wanna attract Pro users
so i think it's a business plan
Yes probably, but I mean lets see what can be done, do you have ultra?
i think it's not like hey you can use Veo3 too enjoy your time
yea i have it
but i consumed all credits in just a week
yea unfortunately, most of the videos had poor quality
yea true
Veo3 is the best if you are trying the selfie style shots.
but things get worse if you try other style shots in general
I think that's why Veo3 videos are usually selfie style shots
but the audio sync is truly brilliant. it's like... how is this possible
Honestly I think that's a skill issue
It can be a skill issue but also AI model capacity
I would love a affiliate program or something where the deep thinkers can tinker lol. i would be glad to share any findings or help make it better if i can. i am currently pushing pass its limitations and it seems to be working
You know this AI model is not a model that can do things properly always.
Lmao I know I'm messing with u
ok lol
Honestly just make stupid stuff and it'll blow up bigger than anything realistic
Make more money from the broken videos
that can be a strategy yes
Brain rot activity is peaking
U can make an ad with a major AI artifact that makes it funny and get more impressions
yea true maybe i gotta try stupid stuff too hm
check my dm
how can it happen only non followers lol
does instagram reels pay ad revenue like youtube?
i heard instagram is pretty doo doo, u need like 10k before monetizing
even then i think its pretty bad
its all about learning what platform has a niche that hasnt been AI brain rotted yet
Still tho, I am sure you could sell them the "VEO 3 course" hahaha
AI vids going pretty big on TikTok rn
there's like 20 big foot vlog accounts and they all get a million views
Honestly, the first time im having success with tiktok is with these AI videos
those arnt viral tho
100k on one of my videos lol, and 2 other ones at 60/80k which is not bad
its kinda hard to do it if ur starting with no followers
But heard tiktok monetization is horrible so idk
phew
randoms who like bikes at least hahaha
Bro sent u a dm, when u got time would appreciate ur help im stuck with a prompt u might be better than me for sure with this
but how do you monetize it? views don't matter tbh
if you can monetize it is matter
cuz we all pay money to generate videos
I mean, you could for example make an instagram or whatever all about "Create with VEO 3" or something like that, and then sell people a course teaching them how to use VEO 3 properly etc
but it's not from videos you made bro
I think Tiktok or Insta cannot beat Youtube for this
Or you could create a video production company on instagram lets say and get companies to pay you for marketing vids
There is a million opportunities tbh, I wouldnt waste my time with Youtube
be honest bro
?
I'm using VEO 3 currently for my own business, ecom and I think it could work out fire
You cannot make profits from videos on Tiktok and Insta right?
From views no, from people yes
Sorry to break the flow of conversation here friends, but I am after what maybe a simple question to asnwer.
I want to create a video with me as the likeness, ie, upload a photo of me and have it create a video of me saying specific phrases. I used the Frames to video function but it starts off with the picture I uploaded, which I do not want, I just want the content beyond that, but the transition makes it impossible to even edit out.
I am also getting no speech, but I see alot of people complaining about this at present. Sorry I am very new.
What you are trying to do is not for Veo3 currently.
You need to look at Runway
then You will have to use lip-sync feature there
Is runway a part of flow?
It's a totally different platform
I don't think it's appropriate to talk about it here
Appreciate your help 🙂
I need unlimited plan to come out soon 😁
anyway to use veo3 for free?
it would be really expensive :0
i heard There's a student plan
#flow-tips-n-tricks・✏️ A 45 year old Vietnamese male dentist is in a dental clinic with a green and blue background, wearing glasses and a white blouse. The dentist says: "Tôi là AI do Kim Cương tạo ra, trông tôi có giống nha sĩ không hehe".
create a video with some information like : A 45 year old Vietnamese male dentist is in a dental clinic with a green and blue background, wearing glasses and a white blouse. The dentist says: "Tôi là AI do Kim Cương tạo ra, trông tôi có giống nha sĩ không hehe".
What happened
Veo-3 fast, 2 generations in a row - no sound! 😩
Anyone wanna help me out
“A tense, dark comedy scene: A cop walks in, a person says ‘Here’s your black coffee.’ The cop suddenly pulls out a gun, shoots the coffee cup repeatedly. The coffee explodes. The cop yells ‘Stop resisting!’ to the destroyed cup. Cinematic, exaggerated, 8 seconds.”
A tense, dark comedy scene: A cop walks in, a person says ‘Here’s your black coffee.’ The cop suddenly pulls out a gun, shoots the coffee cup repeatedly. The coffee explodes. The cop yells ‘Stop resisting!’ to the destroyed cup. Cinematic, exaggerated, 8 seconds.”
why does every image that’s uploaded violate guidelines
did somebody got a fix on the issue where flow tells you "you must be an ai subscriber from google to create" eventhough you are a subscriber?
If you click the redirect button to the subscription page it even tells you you are already a subscriber
Ji
When trying to dub in another language. Do I upload the image or the video ? For heygen
I cannot have fun with veo3 anymore now. consumed all my credits
Quick question, how did you get to be a 9:16 aspect ratio?
abuse the ai till it listens
i use keywords like portrait, recorded on mobile phone, 9:16
doesnt always work
hahaha brooo whats with the jump scare mayyynn
its my payment
ok thankyou for the reply, i appreciate it
hey bro, When I type “speak” in the me prompt, it only moves my mouth but no sound, what's wrong?
Anyone have any advice on how to consistently generate people speaking? I've had luck with adults but it seems having kids speak alway sends up blank. Ive prompted it with something like Child says: Blah blah blah. Dont know if there is any additional tips I can use or if VEO just doesn't have a kid voice lol
Made entirely with Gen AI.
To highlight a few tools:
☑️ Character Voices: The New ElevenLabs v3 Alpha
☑️ Script: Me + ChatGPT
☑️ Images: Midjourney v7
☑️ Video: Kling 2.1 Master, Veo 3, Higgsfield
☑️ Lip Sync: Heygen Avatar IV
☑️ Upscale: Magnific and TopazLabs
☑️ Editing: Adobe Premiere Pro
I locked down this we...
"Create a dynamic and engaging promotional script for the 2015 Dacia Logan. Highlight its key features such as fuel efficiency, spacious interior, modern design, and affordability. The tone should be confident, friendly, and appealing to everyday drivers and families. Include visual suggestions, voice-over cues, and a catchy slogan at the end. Target audience: budget-conscious drivers looking for a reliable everyday car."
Text to video?
how do people generate so much. everything is so limited. i feel like even with ultra plan you can’t create much
It works really nice swifting scenes in the same context, still not able to keep the same character with audio (without using extend function) , any tips?
Is there a big difference between Veo 3 (fast) and Veo 3 (Quality)?
Nope. The NEW Veo3 fast with audio is pretty nice and for just 20 credits a generation. I just wish it had image prompting, but even the highest tier Veo3 doesn't have that yet.
Hey guys sorry for my bad English I don't use it so much but I am from Belgium and I would like to use flow so I am using a VPN but when I try to use it it says that I need to have a subscription(I have it already) and when I click at the message it sends me to a site that says that I have it already but when I come back it's still the same, do someone have a idea what the problem is.
Thx for reading
In fact, I was planning to cancel my Ultra Subscripton when my first month runs out next week, but now the $125 seems worth it for 12,500 credits when I can make 625 Veo 3 fast generations with it. Audio and dialogue are still hit or miss - but at just 20 pops a generation its easier to tolerate the misses.
625 generations is really nice
I just wish they'd let my unused credits carry over to the next month.
Hello, I bought the Premium version of Google, at €100 per month but when I go to flow, it tells me that I have zero credit, I don't understand why when I paid €100
how do you use this program?
Did u buy Google AI Pro?
yo guys how much of differemce in quality is veo 3 fast and veo 3 quality?
i tried to make a stormtrooper video but it wouldnt take it or it wasnt what i wanted idk if im doing it right i followed what the tiktok video said to do but it didnt help
wdym u followed the tiktok video
oh
dk if i did it right
u wanna make videos to post on tiktok?
They say they've fixed the audio disappearing in extend? I haven't tried that yet. I'm too busy having fun with Veo 3 fast. With the price drop I can afford to make more than 1 at a time! I boosted my output to 2 at a time. And I can go back and try to fix some of my fails.
yes
For my purposes - not much! Hard to justify 1 quality vs 5 fasts when you don't know if the "quality" version will be a fail.
how many followers do u have?
if i am able to do it correctly id start with zero lol
on frames to video whats the 2 different image things
where Can I get veo 2 fast unlimited?
oh
Ok I just tried to extend a Veo 3 fast video with audio. It switched me over to Veo 2 and although the Veo 3 audio was unchanged, there was no audio in the extension. :< I won't be doing that again!
I really don’t notice much of a difference either
possible to share the prompt please? it is ok if not, i understand.... every time I do a action scene like this, my characters are acting like amateur actors, waiting for their cue to move or throw a punch lol
Hello, I am subscribed to Google AI Pro but i can't seem to find where i access Google Veo 3. Could someone tell me where i access it?
flow.google
or you can use Veo 3 Fast in the Gemini app
hmm, i cant seem to find the option anywhere. could it be because i am located in the Netherlands?
Thanks for your help i found my problem, it is indeed my location.
hello guys, I'm new here and I need some help please, is it normal to flow take more than 30 minutes to extend a video?
Hello im looking to use flow but it says it's not supported in my country, if i were to buy the Google AI Subscription, would i then be able to use it?
hello just a question did Flow fix the problem of when u extend the video it dosent create no sounds
anyone ?
@ripe path do you know if image to video will get audio soon?
I’ve got to say, I’m impressed. As someone who works on films for fun—like using green screen and doing all the editing—Veo 3 has really helped my workflow.
I’ve noticed a few things that don’t work, though. For example, putting pictures from MidJourney doesn’t really work well for sound. It feels like you have to make the video in Veo 3 first before you can use frame-to-video and ingredients for sound.
I’ve always wished I could work for Hollywood. I guess in a way, I kind of am—just with the freedom to make movies the way I want.
One thing I also discovered is that Veo 3 seems to dislike the word “doll.” It’s really hard to put pictures in there if you’re trying to make a doll-themed movie. I tried it for my Halloween movie, but it didn’t seem to work.
Other than that—great job. I’m happy with the results.
loving the new changelog ❤️
If they make Veo3 fast 10 credits it will be Christmas in June... 🙂
Just wait for Veo 4... and Veo 4 fast... it's only a matter of time until it'll be as cheap as generating a regular image
Once we get Veo at 1080, 10+ secs, extensions, consistant audio, all at a cheaper price than today - 👍
i keep on getting captions to my video but i just want my characterto speak. how to fix this
good day guys
it seems like its not possible currently
been asking this to people and they said it's random
how many videos can I make per day with the ultra subscription
I think if you use all Veo 3 and not veo 2 it like 4 a day. But if you want to use all your credits you can make a lot in one day lol. How ever you want to do it. that is up to you.
Hello its possible videos 9:16 in flow?
not sure have not try it. You could use edit software like Resolve or capcut.
okok
"Gemini Pro members get three Veo 3 Fast generations per day with daily refreshes (instead of a 1-time, 10-pack)"
💯
i wanna play with veo 3 fast but, I'm runnin out of ideas lol
Yeah I just wanna get on the Bigfoot AI videos trend for tiktok and Instagram and need to make a ton everyday
So if I get the ultra I’d hope to make a ton of videos in a day
This looks so fun
I don t manage to have the sound even once on FLOW with VEO 3 quality or fast. I am very accurate on the prompt , i even make write the prompt by gemini or chat gpt, but nothing works, no sound.
A solution ?
I'm having massive difficulties getting veo3 to make the videos I want. They are nothing fancy, not excessive it the amount of detail and actions, on the contrary, what I'm asking is quite simple. It just looks to me that either the prompt is ignored almost completely or the videos are "dragged out" almost like the 8 seconds are to short for what I'm asking (even tho it's absolutely enough for the actions to happen). I'm not sure if it's because of my prompts (I tried self written ones, chatgpt generated and gemini generated, all for veo3) or if it's the capabilities/lack of capabilities of veo3. Thoughts? Thanks
Instead of using " " for the lines use something like the character then said: good morning. By using : you won't get the subtitles most of the time
I made a video added sound cues in prompt and got no sound
that happens quite often to me as well
That’s like stealing because they still charge you the 100 credits, now your stuck with a soundless clip you can’t use🤦🏾♂️
but if you a period at the end of the sentence, at which point the AI will know that the person is still talking, or you are describing the next action?
but if you put a period
like.... he said: good morning. she smiled and she replied: good morning to you too!
hello i'm newbie here. how start flow ?
Can anybody help me?
That's the limitation. Object consistency or sound but not both. The list of which method has what is in their faq.
Hello, idk if I can post this here, but all the videos that I generate has a 8 seconds cap, how do I do longer videos?
Currently you can't generate clips longer than 8 seconds
i cant get audio to generate in any of my flow clips. does anyone have a suggestion for another tool that could fill in audio to my clips
there is a option for it where it says veo 2 with audio (Beta)
sick, just found that. thank you
welcome
im guessing it doubles the credit usage when selecting audio right?
just went from using 10 credits per output to 20 now
it is not Veo2 with Audio, but rather Veo3-with Fast Generation via Text2Vid , which cost 20units... which is an improvement, since before you can only use Veo3 if you spend 100 units
ohhh
yea different models different credits
Great announcement today! Veo3 fast 20 credits.
yall why when I do frames to video it doesn't have any audio? is there any way around this where I can retain the previous frame?
wdym do u mean like u have to edit it in urself?
Veo 3 with a provided image or clip can only generate video with sfx, not sfx + voice
So make sure to emphasize that there isn't any talking
You can do crying, screaming, other sounds except talking
then it will generate with audio included?
Not always but higher chance if you make it obvious that it's just sfx and not talking
darn that's a pretty big limitation, thanks for the insight
Has to be Veo 3 quality though, u aren't able to do Veo 3 fast with provided image/clip
If you use text to video, then you can do all the talking you want
this might be random but were u ever a roblox developer?
I was all over then I became hated 😭
do u remember overcastan? I was like 13 back then
I used to use scripting helpers a lot
for sure, what a coincidence
is flow is free?
If you pay first then yeah
JK I think flow is, it's the AI credits and subscription that gives access to the model idk
do u know if it's possible to do the free version over and over using new emails? its like 1000 credits for the first one free
Still without audio for you guys too?
I have audio
anyone help me to understand this:
do u know if it's possible to do the free version over and over using new emails? its like 1000 credits for the first one free
this for me:
Mf I was trying to give u an answer😆
yes I believe so
how this work man:
do u know if it's possible to do the free version over and over using new emails? its like 1000 credits for the first one free
how i do it please help?
can you please use chatgpt to translate my mesages into your native language, then type to chatgpt in your language and ask for an english translation then send that to me
Cause idk what u want
I am just asking about, Is there any methode of using flow for free? yes or not
doesn't look like there is in your country. Try using a united states VPN, create a new google account, then retry getting a subscription
ok
is this able to genrate movie from secript ? and what maximum length it can genrate video?
yep! thats showes free in usa
oh cool
it generates 8 seconds of video. With 1000 credits you should be able to generate 80 seconds total with Veo 3
from one account?
or is there any thing like invideo.ai for free
😔
kling ai
in one account
nop
artlist also good for temp gmail accounts
brother i need for youtube video from secript that genrate visuals for whole secript ,
it is very difficult first genrate secript than genrate prompt for each scenes and than genrate video and than creat like multiple account an dat last get no views get boring?
i am doing same as you
in artlist it takes like 20second to swtich an account
and there is no queue either
use whisk for 100 generations
per account
anyone help me to verify usa number for create account for flow?
Veo3 ain’t free you need to pay
in usa it showes 0 for one month
thanks, i also found footage there
i cannot get audio for text to video in veo 3 to work, tried over 10 prompts and used gemini to engineer. it just does not work.
Generating it on something like runway is working but that is so expensive
Weird how a third party api service is more consistent than googles own flow. makes me think its a flow issue and not a veo issue
there is option for veo with audio
Ya trying that on Flow
but the audio fails most of the time
Also are you telugu haha
no js bahubali fan
How does the student thing work?
is pakistani student get it free
thats only for usa rn
you can use vpn to access that
u need a student email for thyat
student from usa also?
Ok
which sites you use that give all prompt from scene of youtube secript beacuse chatgpt and other never give complete or not even consistence
i use chat gpt. you just train ur ai with every correction till it gives best results
i have train my gpt very well
can you share your work link
can any1 help me out. We've heard all about Veo3's lip syncing ability. But i have a video with a bunch of ppl singing along with a track. is it possible to use veo3 to remove that lipsyncing.
if its possible, how should one do so
Hey all! Flow noob here. My country does not have access to Flow yet so what's the best way to experiment with the platform?
u dont need audio?
use vpn
Thanks! I'll try this out.
ok, i have a track. There is a jingle playing, and the ppl in the video are currently singing along with that jingle. The audio isnt a problem. I just need the people to stop moving their mouth in the video. As in, they need to have a closed mouth, serious face instead of their singing faces.
dang
thats a hard problem ngl
try using prompt "not singing" or try using image to video
i dont want to generate a new video. I have a video. I just need them to stop singing.
is there even a way to upload and edit videos on veo 3
???
I am making a scene in a movie where a mother is saving her kids from a flood and even though the kids aren't talking I can't get any audio to Gen
I'm making a short film
I think the kids crying might be an issue tool that out we'll see
Not gonna lie actually turning my 93 page screenplay into a full length feature film I have distribution network planned I have already submitted my script to many writing competitions
Have the first 40 minutes of prompting done before I started generating
Hey guys I can't use Flow cause it looks like I'm not subscribed, but I'm actually am, how can I fix this?
It s absolutely unbeleivable. With gemini VEO3 works perfectly, you have sound all the time, and with flow which needs ultra at 124USD it works one time on 10
I am a newbie, how to create a prompt that sticks to the story as closely as possible to create a video
is there a way to keep the same looking character throughout a few clips?
same here
Hello! In Flow, is it correct that image to video is not yet possible with Veo3? And after adding Veo3 api allowlist, how long does it usually take to get allowed???
use same seed
use deepseek or chat gpt to create ur prompt
veo 3 is still in beta so there is limited editing
hey guys i need help, i have a pro subscription on my google one account and when i try accesing flow with this same account it says i need the subscription with a subscribe button on the side, once i click the button it says that i have already got the subscription but when i open up flow agin nothing changes and i still cant access it, any solutions?
veo 3 is definitely worth getting into early and becoming good at prompting
ur country?
Guys unsubscribe this sh..t, and take pollo.ai, you have character consistency, ingredients, voice and sound everytime. yuo can start again the scene from where it stopped
And it uses VEO 3
flow is just so bad
veo 3 works perfectly, it s flow not working, try it on pollo ai or even gemini it s perfect
u for sure about it using veo 3?
gemini is good with veo3 , you have sound, voices all the time. but you can t extend or jump to, that s why i took flow. but you have sound 1 time on 10, it doesn t work, they have to be ashamed to make us pay for their beta. But pollo ai does all that with veo3
yes i m doing it right now.
you have all the models, veo2, veo3 with sound, klig ai etc... you have at least 20 models, but you choose veo3, you generate your video from texte or image. If you want to character to be consistant you can continue the scen taking the photo of the last picture of your video and continue,
pollo ai is expensive
but it worksflow does not work
gemini yes
yeah i can do that on flow guy
but no ingredients, image tc
for 1/3 the cost
how ?
just like u said
"taking the photo of the last picture of your video and continue,"
u can do that on flow
you ask it for sound, you don t have already most of the time
yeah its called learning how to use the model
ur paying a premium for someone to do the main work that you can easily do for free
yes theoritically, that s why i took it. but in the end 25 videos; 0 sound
yeah because u prompted it wrong
ok so my prompt is good in gemini, and poolo, but wrong on flow ?
can i test your prompt.
it s a goo exercise for me, m here to learn
idk ive never used gemini and i get great results on flow other than UI bugs
i want a man in his car with his girlfriend, saying just: Best experience ever!
i also wanna learn
i want a man in his car with his girlfriend, saying just: Best experience ever!
Can you prompt this
for me, just once please to see
gimme couple seconds
thanks
im from the US it worked on another account for me but for this one it doesnt
the guy is saying that to the girl for example, just for a test
i have no idea then its just a bug. you go to customer care they will solve it quickly
its probably a blocker to prevent people from buying accounts for people without access natively
oouu
@umbral root does that line "best experience ever" work on other platforms
yes gemini and pollo ai
interestingggg
we can chane it if you want
It's the best experience possible!
or what ever
does it look good?
gooner
the sound is amazing as usual lol
ima try something real quick
i might know why
in gemini how many credits does it cost
gemini uses fast veo 3
for me it cost 20
in flow not in gemini
i understand, im asking when @umbral root generates that best experience ever video, how much does he pay
in flow now you have fast veo 3 with sound and veo3 quality with sound, i do both anyway
i don t see the cost on gemini
okay so theres a good chance ur using quality veo 3
after a while he says i just have a few video left
i can t choose on gemini, it s written video and when you go on video it veo 3.
I use veo 3 quality on flox as well
i always use quality, i switched ones on fast to be sure. But i m always on quality
but u said in flow u had the issue of it not working
A guy is presenting a podcast in the USA, ans he ask to his guest the exact following words in english: "what are your sources of inspiration?". The video has to be edited with the sound of the room, and the voice of the characters. We have to hear exactly the sound of his voice saying the exact words of the qoted sentence of the prompt
dude im not wasting my stuff to prove you wrong when i just solved the problem
figure out the rest or pay extra for someone else to do it for you
A guy during a podcast ask to his guest the exact following words with a perfect lipsync in english with a normal english accent: what are your sources of inspiration?
ur adding useless key words
are these on ur pollo platform?
because i try anything
no on pollo platform it s this
flow? when did u generate those and were u using quality?
prompt is trash but it proves the point, after switching to quality it generates with audio every time
👎
yes i know, but my goal is to make work veo3 on flow , what i don t manage
its like learning a programming language
yes io generate i quality
idk man honestly
the gemini chat probably elaborates on your prompt thats basic, just like im sure pollo does
flow seems to be the raw interface
maybe flow generates its own prompt for your prompt idk
yes , i d be happy if it was just a prompt issue, but what prompt did you make ?
if you put the same prompt as me and it doesn t work to me, the issue is flow
i didnt use the same prompt <3
A 9:16 portrait video from the dashboard view of a car driving home at night, faint carnival lights fading in the distance through the rear window. Inside, the soft hum of the road and distant fair sounds still linger—muffled laughter, a faint carousel tune, and the low rumble of tires on asphalt. The cabin is dimly lit by passing streetlights and the occasional flash from roadside signs.
The man is driving, one hand on the wheel, calm and content. His girlfriend, sitting in the passenger seat, leans back with her head slightly turned toward him, a cotton candy stick in her lap and a small stuffed animal from a fair game in her arms. With a wide, dreamy smile, only she says:
"Best experience ever!"
don't add subtitles.
im pretty sure it explains somewhere on flow how to properly format the prompt
chatgpt does it for you anyways
yes, so you just said > Only she says: "blabla"
?
in you prompt, to make her speak, you said, SHE SAYS: "best experience ever", nothing magical. All the rest was the description of the scene. But it s just the sound not working for me
adding extra details about actual SFX rather than just VOICE will help it generate audio
theyre 2 different things for the audio feature
you have to quote it and be specific about whos talking
adding details about noises unrelated to the voices helps.
plus you didnt word that properly in english, you basically worded it as if shes talking to herself lol
https://youtu.be/DtUdlODKWWM?si=qPtA58j0IhJYllL4 - for those that always wondered what consultant is like
All the bullshit you encounter as a consultant. Based on a true story. This is just the beginning.
#consulting #consultants #consultantlife #corporate #corporatelife
yes i quoted it everytime, and then i stopped it and it worked once. But the fact to describe the sound around is maybe the key to trigger the voice
yeah SFX and Talking are 2 different things from what it seems
like when extending clips, people get confused to why they cant extend it with audio working and its because they can only use SFX
what do you mean, when you extend you can have sfx but not Voices ?
Text to Video, with speech and environmental sound.
First Frame to Video, with environmental sound.
alright. and image to video ?
honestly idk too much about the extend or jump to features, but i think they use the frame to video stuff
image to video is "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.
so it only works with jump to, not extend
i descrtibed the environment , the sounds mostly and it triggered the voices
