#flow-general・🌊
1 messages · Page 29 of 1
let me check
it was using veo 2, now its veo 3
Before couldn’t you Just take the last frame and animate it like any other image to video?
Or extend takes into account of everything
jump and extend always there since day 1
Okay just used the insert feature that’s sickkkkkkkk
They’re definitely using nano banana for that backend
cant use it yet
What
🔥🔥🔥🔥🔥
so veo 3.0 is officialy "deprecated" according to the flowvideoconfigmodel
Let’s see that video! @analog mantle
Wonder why Veo 2 is even still there
a racing car in the desert , then hardcut to cockpit , then to driver face
yo thats actually really good
Veo 3.1 is on flow already
yes
yea we're playng with it
go try it guys
Looks like it can’t generate 30s/1m vids 🙁
I tried a found footage horror film and I received an error "failed to generate." 😅
Insert is actually really crazy asf
hover your mouse on any generated video in project page, u should see the "pencil icon" button
It’s like video to vjdeo but you only add can’t take away
thanks
i just try the extend on scenebuilder and wow , no janky transition..so smooth
I tried something different and another failed 😂 😂 😂 😂 lool
Gotta test it out some more. Apparently you can insert things to existing videos with “insert”
not supported by vertical mode as yet
No need vpn
no need vpn
itll return
The band Toto had a song back in the day. Simply called 99. Anyway, come on. I think that's gonna be the anthem for flow. What's with this 99%?
Music video by Toto performing 99. (C) 1979 Columbia Records, a division of Sony Music Entertainment
Maybe
I mean honestly come out. That makes sense but it's like no communication at all about that.
i use vpn korean to get it ahhaa
insert add thing on the video
Honestly, the The fact that it's over capacity or can generate a video is not really an issue. I get it. However, it's the problem that it just freezes at 99% and then doesn't tell you it failed. That's an annoying waste of time. It should immediately tell you if it failed.
a cooking scene , then hardcut to the chef hand , then to cooking pot , then to food
edit , add thing etc etc
So if i have a full body scene i want to that moment to be in a close up i can edit it
let me test it
That would be extend
Insert adds objects to the video
Hi, guys, so what, there is now extension till 60 second and any features all had waited for?
It’s extended in builder?
Man all my videos with veo 3.1 just keep getting stuck at 99 percent
yes
wow. the graphic 3.1 amazing but still 8 seconds
Yeah
i mean im used to it happening lol. but since the veo 3.1 update its happening 100 percent of the time
The 99th thing is super annoying.
Ninety-nine percent is now happening. 100% period. That sounds like a really Effective political campaign slogan.
But yes you're totally right Ninety-nine percent 100% of the time.
99 % zzzz
So by the way, the percentage if you guys don't know is actually just a timer, it's just an animation, it's not actually a real percentage. So when it reaches 99 and it holds it basically just means that the video is not in generated in the time it's taken the timer to get to 99%. That's all it means so. Video is not generated for whatever reason, probably because of server load.
Wasn't it John Wooden out of UCLA who said we never lost a game, we simply ran out of time before we could win?
veo3 down in gemini? I keep getting "I couldn't do that because I'm getting a lot of requests right now. Please try again later."
The server load is heavy all over the place and I'm not surprised. I think everyone's experiencing this but Gemini is the only place where it's explicitly actually saying
quality works, but fast is stuck
veo 3.1 is out probably getting spammed hard
its working ok for me
finally FFLF
whats FFLF?
ffirst frame and last frame for veo 3
I didnt see any official announcement about the new features, all of it was expectations.
guys do extend and ingredients to video work on veo3 ?
is it better than veo3 qualitywise?
extend uses veo 3 yes
I already test for new features now support audio for function ingredients it's very helpful for management characters
if u pick veo 3.1 quality to extend a video, apaprently it cost 100 credit
u can use veo 3 model now for ingrediesnt to video
Hello Im new here and have a question concerning testing flow, I wanted to test the functionality of video gen, but could not create as it is saying I need more credits and it only offers to directly sign up for a monthly Ultra subscription. IIs there a way to get initial test credits before buying a license?
yeah you can just get the one month trial for pro plan
@apoke111 where are you read this one
🤐
where do i get this, I cant find that
Thanks a lot
seems the same to me, still take your input images and make the video have alot of contrast and kinda cartooney
lol acting like the network tab is a confidential informant
To be fair, for 99% of users, it is
Is it working now?
What the hell, u sure without sound?
I never really got ingredients to video concept
My silent mode was on, sound works
iv managed to get 2 clips using 3.1 fast with ingredients, but tbh, it didnt impress me at all, didnt seem much better than veo2 fast. but maybe i just need to improvide the prompt abit. or the servers overloaded. also, yo ucnat use ingredients with 3.1 quality, which is a bummer.
rn
where? can we watch ?
That's a fantastic news 🙌
anyone got a fix yet for the 99% issue?
99% issue
Welp there it is
scene builder is working fine
I wonder if veo 3.1 fast will still be 0 credits for ultra. it shows as 0 credits for me but ive seen other ultra users showing 20 credits
How can I get scene builder on a mobile device?
im on pro and its 20 credits, and 20 for quality.
100 for quality
yea
mine shows 20
How good is it
stucks at 99% everywhere
I mean based on that wok clip, it’s basically.. Veo2? Like how did anyone at Google think that was a game changing clip. The voices sound a ton better, not like a robot with a sore throat talking into a shoe box so that’s good I guess..
Ahah
Yes it still is free for ultra users
google tryna get back their investors after sora 2
yea yea:)
How much better is it from 3
I’m extremely annoyed I ponied up for ultra for 3 months at launch of veo3 and it felt like an alpha
it looks like video generating got bananas 🍌
@flowbygoogle just confirmed to me on twitter that 3.1 fast remains unlimited for ultra. yay
Anyone else running into upscaling lag time? After multiple page refreshes, upscaling/export is just spinning and spinning
That’s been a known bug since launch.. it works on and off, better to do it locally
I really thought they would’ve have 1080 native by now
i bet loads of people are generating right now, mine was doing this too,
Is google still offering a 3 month half off trial on ultra?
i wish they would incorporate a Queue or something, even if it took 10 minutes to get to my generation, at least I could Queue up a few things and then come back to them once they were done. Not even knowing if its working or not is prob putting even more strain on their system because people just keep re-submitting the same thing cause it doesn't look like it took
I was really expecting some quota upgrades with veo 3.1 compared to what Sora 2 provides
wow 1 minute video true
I mean sora isn’t having an issue with it
Extend forgets the audio is there a way to keep the character sound same
On what? lol
u did 1 min video?
extend
why does 3.1 animations look like veo2? veo3 was better than this :/
extend is working fine
😂
same here
im in the middle of creating a music video 🙁
yeah, gonna be back to 2am gens just to get it to output
i cant do that upfor work at 5am lol
iv go untill around 3pm friday afternoon to get this video done, hopfully people get bored and stop gerating anything and everything lol
For how long can we create each video in VEO3.1?
forever
Not for me
Reaaalllyyy???
Like, can I create a video with 2 / 3 minutes?
1-2 minute
If you keep using the extend feature yes
whats the drawback?
that`s a joke bro
Got it
but for each video, how many seconds do we have?
What I don’t understand is that it just takes the last frame and animates it what stopped anyone from using it before Veo 3.1 dropped and plugging it in the first frame of Veo 3 of the last frame of your video? Whats didferent?
I've started in VEO3 a month ago and I'm learning...
im too
so veo 3.1 better than sora 2 pro
am I the only one ? 
no
its hard lagging right now
No, it's definitel not working right now
Hard to tell. 2 completely different style of output it
Okay I should sleep
Hi sock, so is it confirmed?
❤️
Thanks buddy

so does this mean veo 4 is delayed?
5 months to release .1 update. we will waiting 2 years for veo 4
😂
Nahh buddy...
Google is doing very well!!
they pushed this update only cuz of sora 2
Lol. VEO is adding features to its current model, not coming out with a completely new model yet
I did the same to "trick it" into extending the scene with VEO 3, in Scenebuilder it lets you save the last frame as an asset and then you can use that. Problem is, since it's not technically the extend functionality, if you have smoke or moving water or other moving elements (like clothing blowing in the wind) it will not keep moving at the same speed and can be quite obvious where the seams are, even if you do a bunch in video editor to make it look the best.
sora 2 is still sota. veo has this plastic feel.
only played around for a lil while, but the updates to Frame to video and especially Ingredients to video are HUGE improvements from yesterday
Start/end frame will be huge in making extended scenes and videos
Sora is too twitchy and spastic. Even when I prompt it to not be.
was able to achieve 3 shots with the new model that took a week to do in previous set up
Sora 2 is literally VINE. It feels like the entire model was trained with all of vine
What you mean
Haha yeah it feel vine tiktok coded
modelwise is 3.1 looking better or sora 2?
does flow allow generations in spanish with veo3??
Runway-3? Who else compares it? Even Runway Gen-4 has stopped comparing it.
from what i managed to do, it didnt look much better than VEO2 tbh, but could be because everyone is using it
i think it depends for what are you going to use Sora 2 or Veo 3.1, Sora 2 is more tiktok/edgy things, Veo 3.1 is more "Professional" Looking
I use both, depending on the project
i take back what iv just said, iv manaed to get some more 3.1 clips, its way better than veo2 for movement, and 3. and the use of ingredients is great
thats from official veo 3.1 benchmark by google.
Frustrating to be waiting for my Pro credits to reset when I want to try 3.1 (:
I upgraded to Pro but I can only generate 3 videos through Gemini and zero through Flow? Am I supposed to have access to Flow? I tried asking support but they thought I was asking about watermarks. Can someone tell me what I'm paying for and how to access it?
people never really use text to video so from image to video veo wins
you can check how many credits you have in flow by clciking your name icon in the top right hand corner
where? can you send me link? i haven't used any credits.
hmm when i am logged into flow i see this in the top right when you click the blue icon it then shows credits, are you definitely logged in? or refresh the page or try different browser maybe?
I was able to log in and see my profile in the top right on safari. wonder why i can't log in on chrome
@main saddle So if the APIs are already updated for freepik, fal.ai, higgsfield…, when will the 3.1 models be live on the Vertex platform ?
@main saddle why you delete the x, it was cool
also.. all the updates are insane!
Not so great for Swedish language and text can someone please look in to it
i'm in france ( paris ) , and my problem is that all my avatar have canadien accent . it's super annoying . #help
Huh, I'm finding degraded quality on my generations with 3.1 fast. Anyone else or am I just getting some unlucky gens?
me too , the quality has decrease
I got excited to try some extends but I noticed the audio hiccups when it tries, especially with music.
Retrying old prompts has gotten some WILDLY different outputs that look much worse. 🙁
Im turning back to 3.1 Quality ( beta audio)
Not at all
1- i just got to check one extend. it is bad. it uses not the context of the video it seems like kling. it uses the last frame which is bad. an unrefined, out of context, last frame. same as doing it manual. 2-facial expressions totally distort the subject 2-very buggy, like veo 3 on launch, is like the old bugs came back , like subtitling of off-voices . 4-microexpressions distort the faces in establishing static shots as you were to create magical effects on kling at suff speed. this is unbelvable. this is just veo 3 with start and end frame. lagging in fluidity against sora 2 and wan 2.5 already. what a fix can do?
stiff as runway, with voices and distortion. same veo 3 like before. i dont see improvements.
oh and the pricing, is etill extremlye expensive
If you do a quality 3.1 I've noticed better results with extend, but still not great. It seems like it took 1 second of video or something because I can hear and see a glitch of the sound with the extended clip that's sampling the first.
My most recent generation still ended up changing the voice of the speaker though.
no. also i get to made my prompt in spansih, first time i do that. i m from argentina. speak castellano, which is a variant of spanish and on top of that, soutamerican voceo. but i prompted it in neutral spanish. the argentinain accent when prompted in english, no problem, with spanish promptng it mixed subtitles, voices in spoansih and portuguese. so the promptng is not quite the same. the encoder may take the tokens with other meaning. which is somethign to be expected. thing is....this is google
(i mean with no, i didnt get any better with quality. sorry it sounded too stoic RiskyWhiskey in the preivous message)
you got distortions or what specifically? just to try to avoid any possible problem that may arise due to this update
I posted some examles in the flow gallery of what I'm experiencing. It feels like there must be some sort of bug with the release right now.
I ran through the project, which had problems with dialogues in VEO-3.
VEO-3.1 - no changes. The same sound distortions, dialogue failures if there is more than one person in the frame, and the general plan.
But $140 for unlimited access is still... the closest unlimited access is Runway ($90), but the quality...
im gonna check out later then, thanks!
Hi everyone, question
Now with Veo 3.1, what is each video length top...?
8 seconds I think
thanks, I thought there were a change
Everything I've made, new or extended, has been 8 seconds. I'm gonna wait a day or two before I do more generations since the website is struggling and I think there are some bugs.
Still no 1080p
Regarding the new announcement, what's the incentive for "Editing your videos directly in Flow" compared to just adding the object in Whisk?
Essentially, creating a better first frame without the risk of wasting your credits on video generation
8 seconds, Still no 1080p
I need to make a two minute talking head video. I have an image of the person. Would Flow be good for this you think?
It would be very expensive. If it is just a talking head Flow wouldn't be a very good fit and you'd likely get consistency errors.
bro i can' t get away with the canadien accent on my videos. they need to fix it quickly.
I’ve only tried Sora 2 so far — looks amazing for short clips, but I’ve heard Modelwise 3.1 has strong consistency too. What do you think?
it's confirmed that there's been a bug since last night. So i wasted all my credits , at least 1000 credits in 3 hours because i wasn't getting the result i wanted.
Does extend work? It doesnt work for me
Veo 3.1 is a slight increase from veo 3 but sora 2 is still way better in naturalism
Sora is for social media Veo is for actual films
thanks sherlock
How can it be better in naturalism lol
sora 2 looks way more natural than veo 3.1, u wouldnt know sometimes its ai without the watermark
It looks like someone with a phone takes it sure but you can’t do anything professional with it
sora 2 can make 25 second, veo still stuck at 8 second.
How? Where
sora 2 pro storyboard feature
Ah it’s released?
my cynical take is, they're all pretty bad. but Veo and Sora suck the least. LOL
it's an emerging tech...it'll take time
google prob has sora tier quality internally.
Kml id say next year you won’t be able to telll the difference
it's also a function of power usage and scalibility etc.
ya video generation gonna get saturated early next year
From my perspective, what OpenAI is really good at is marketing — not substance. 🤣
where is google marketing team hiding at
perhaps...Kling was always pretty impressive to me TBH
and the Asian models are really catching up, notable China, etc.
Seed* has been insane
they're all good at their own thing IMO
Klings start to end frame is really good
I’m Korean, but the videos generated by Sora feel strangely unnatural to me.
I agree we just need a generator that’s good at all things
Every generator is good at once or two things but suck in the rest
IMO it's like showing up to build a hosue and having a hammer, a tape measure, a saw, a level, etc. Which one is the best ? Well....the question doesn't even make sense! They are all different. thats how I see AI. they're all tools for different uses.
I am trying to get more consistency in voice when I extend my video. Is anybody else having this issue
Right now with AI it’s best to pare it with good editing for best results
yes
Good old manual editing is still your best skill in this space. Get your hands dirty as the old timers would say
I have many friends in the ent industry, music, gaming,adult, etc and everyone is paying attention (well the ones who are good). But they're going to be using the tools to leverage their goals. not replace them. thats gonna be key. in 2-3 years this stuff is gonna change literally everything. but some will dominate
Yes, I been in the after effects game from 2008
AI is better used when it assist you not a full on replacement
yes - supplement not replace
I struggle with voice consistency on the extension of the first generated clip.
I struggle getting voice consistency when i extend
Not sure if it's OK to mention this on here, sorry if not, but there are tools that can be used to replace the vox and overdub it, if that's what you're saying (ie voice is not the same)
Hello guys
My questions:
Is the edit function also available in the Scenebuilder?
Veo 3.1
Yup
thanks
HI. any tips on keeping voice dialogue consistent when you extend the video? Totally struggling on that
Hello all! I am new to the forum. I just got Google Flow a few days ago and I am trying to do a short horror film with it for Halloween. I am have trouble getting a monster claw to grab someone and pull them off screen. I have used prompts like "the monsters claw grabs them and pulls them off screen / off frame / off camera ext." In every case the claw comes out grabs the character and then does nothing! Any suggestions? Thanks so much!
Flow IOS app wen
Yes. It’s feels like the extension is the old 3.0 model..
Is there a way to get rid of the very contrasted look that veo has?
import it in a editing softrware and lessen the contrast
Should I use my flow credits on Veo 3.1 quality version or fast version?
Is there much difference between the quality and the fast version?
quality version much better in prompt adherence
how do i remove veo water mark in flow?
Looking around online and I’m seeing others are having flow issues. Is this just how 3.1 is or temp glitch?
my same json prompts with veo 3.1 seem less compliant, regular sentences seem to work better. Spelling of words is still very very bad. I have no solution to making test or domain names some out correctly e.g. junkmailmagic.com
lol veo 3.1
Interesting. Having it limited to pro is gonna make sure it works and is slow rollout. Not sure on that time length happening yet though.
i think veo 3.1 can 30 seconds but actually not
veo 3.1 is 8 seconds
Video extend does not maintain character consistency at all if the characters' faces cannot be clearly seen in the final frame of the previous video. Very disappointing.
but why in heygen has different feature
not great with heavily compressed 8 bit color
also its not just boosted contrast, seems like there's really strong sharpening/edge enhancement being done
please remove the free the traffic is so high
I can't even generate a video with 3.1. I'm getting temporary unavailable try again error all the time
you could try adding prompt "low contrast" in your camera direction block. It worked with Veo 3, not sure if it will work now -
Right now, it's 8 seconds at a time, but one day, you can generate an entire short movie based on a completed script. Okay, maybe not you, but your kids most likely. Imagine that.
I keep failing to generate videos. Do you guys experience the same?
same here
Same
Same here. Gets stuck on 99% for a long time then failed to generate.
anyone know if its fixed yet
I've been waiting for 2 hours already
me too
I can't create videos on Flow Veo 3.1 right now, it's slower than usual and when reaching 99%. It says failed to generate. Under pro subscription too. This is so disappointing especially if you're a paying subscriber. ^^"
I agree with you
it definitely is. it would at least be nice to hear why its happening :/
or at least some acknowledgement at minimum
Me too get stuck on 99%
i have an ad to deliver and its getting stuck on 99% 😄 cant finish it
i programmed a python prompt organizer cause i was tired of losing them to these errors 😂
I use Notepad 😄
could you please elaborate on that?
just some example prompts shown
quick easy copy paste. prefer it over having 300 notepad tabs personally
What's wrong with flow all of sudden
Like the veo3.1 isnot making good results compared to morning
i mean most of us cant even get a result at all right now
same , i came to complain about this
Oh my and i have deliveries to submit
Sora not in my country, generated 1 clip on 3.1 and now the 99% error. Going well.
i just bought it to use tonight and cant use it at all lol
any sora2 invite code? was using sora1 but was fast overthrown from other competitors.
yep its broken
why?
is flow working?
u pay ultra plan for error 😭
fck
3 hours now. im going to bed
That's honestly worse to be honest, considering the price and all. 😅
Ugh, hopefully it'll be fixed by tonight. Maybe evening is usually peak time but it's never been this long before. ^^"
There is a message coming up saying "Flow is temporarily unavailable please try again later"
Same here, its happening to everyone, they havent fixed it yet
Flow is temporarily unavailable. Please try again later.
Ah joined here - I see everyone is experiencing the same issue
how can i create a video guyss
Hey everyone!
Could someone please help me out? I’m a bit confused about how to generate videos longer than 8 seconds. As far as I know, the 3.1 update should allow generating videos up to 60 seconds — or am I mistaken?
I'm currently using the scene extender with very clear prompting with a script, but since the update, I'm finding that the generated scene is unusable. It generates it's own random script, completely irrelevent to the scene, doesn't do anything that I'm asking. Camera movement and direction is being completely ignored. Example, I've asked another character to speak with the camera being a close up on their face - The previous character continues to talk but a completely random script, or sometimes, no audio at all (but without the 'no audio generated' warning).
Anyone else having similar issues?
I've not seen anything about 60 seconds. As far as I know it's always been 8 seconds.
I’ve also encountered another issue after the Veo 3.1 update.
As shown in the screenshot, the generated video stops at 99% and then just keeps loading indefinitely.
There’s nothing inappropriate or against the policy in the content — it’s just normal actions and scenes.
thanks bro
If I'm not mistake it's 15 second for normal user and 25 for ultra.
still having the same issue
For the new 3.1 I mean, but i don't know how to make the length longer? Only achieve 8 second for me too.
I'm also facing same issue
Yeah I think this happens like 2-3 hours ago it's still ongoing right now
I have the Ultra subscription, but I still haven’t figured out how or where to generate videos longer than 8 seconds.
I don't even have other platform to make the videos so this suck I shouldn't put my egg on one basket lol 😭
🥲
Seems like it takes forever to be back to normal....😫
Flow is constantly refunding my credits for every single video
I am not sure what's the problem as my videos Re not inappropriate
Anyone got a idea?
It seems like some kind of global error — I can’t do anything either, bro.
🥲
I'm just glad it's not just me going through this
i Thought i only had this problem
Exactly

😂
What projects have you guys been working on so far?
Glad you asked, I am just testing things out so far but I found a decent trick up my sleeve
Complex scenes usually cant be done in a single way so I in a single turn make it do just one pose, take the screen shot from where it went wrong and continue
By the way, regarding the new features — has anyone figured out how to generate 25-second or 15-second videos?
We can?
Isnt it still 8 seconds?
Yo how come Google have server problems ?
Yes, videos perform a lot better when it's frame to video than just text in my observation
Yeah!!
lol they are supposed to "be" the internet
We can extend with flow of course
Oh you should
Facts
@ruby lily when will it be fixed ?
I think i read in #1319424187730755636 they're having resources issue
Theyre trying to get it done
Is that even possible? For Google to run out of resources? Nah, I don’t believe that.
Neither do i
Maybe its all used to train next gemini model
On the note of gemini I recently talked to a data analyst and hes pretty concerned about his job. On his team of 100 about 80 were removed and 20 were kept to supervise the AI
Its crazy how fast the jobs are changing
In dot Com bubble we lost many jobs and multiple new jobs were created with more value. I hope it's same for this one
Did anyone already use the new EDIT feature? Is it suppose tho show next to ingredients or something?
Oh yeah it could be a ai bubble this time, I already use ai for so many of my work and feel captivated but scared of it lol ^^"
I tried yesterday and it surprisingly worked very well, it actually a small pencil icon on the video itself (along with the other small icons)
Not getting any audio using scene extender. Results are really not what I'm asking for either.
thanks ! Want to test that real soon!
Hello guuys! I'm new to this channel.
And I want ask a few questions.
1.All those videos from tiktok where it shows the sora2 logo is generated with Google flow?
-
Can I use Google flow to generate video and audio?
-
Could I monetize this on social media? Tiktok, YouTube
Hey everyone!
Is anyone else having issues with Veo 3.1 today? My video generation keeps getting stuck at 99% and never finishes. Any idea what might be causing this?
It seems like some kind of global error check the above chats same discussions were few minutes back
It is already occurring for a few hours.
Got it, thanks
@high fulcrum when will you fix ?
Wonder if the rollout yesterday didn’t go well. Noticed a lot of bugs.
They should at least allow the option to use the legacy Veo 3 cause so many user already complained the low quality of Veo 3.1. And the bug going rn. 😅
My Veo 3.1 generations work, but I can see them only in My Library.
The video does not play. If I download, I get a .mp4 that I can not play.
Also on some generations, the entire video does not show up in my Flow generations list.
Google flow is same as sora 2?
No google is from google
Sora 2 is from openai
Hello we can only help with moderation issues, please only ping mods if there is an issue that goes against rules-and-guidelines. Thank you!
Erm has it been resolved? is video generation going through now?
No
working now?
Not yet
🙁
Flow is taking ages to generate anything. mostly it fails ? anyone know why it is happening? is it because of veo3.1 ?
yeah its because of 3.1 cant do anything
i tried 22 attempts since 6 hours 23rd attempt worked 24th again failed 🫡
never experienced any of these AI animation companies who had a successful update with a new version, always bugging like crazy
Native 1080p output yet?
I thought I was the only one who couldn't make videos,
its everyone
this is probably the biggest server overloaded ever happened
Why they make it free 😄 ,
My paid subscription is also not working for VEO 3.1 fast 🙁
its okay some folks who never needed it now can try it for free. so we can have problem generating for work while paying for it...
absolutely
My 3.1 sub also does not work
I’ve only succeeded once and the quality was abysmal.
My credits were charged but the video is just a black video😠
I get that you’re offering it free now... but some of us are actually paying for this and using it for work. Please, improve your service. We depend on it.
I can get past the 99%. It's always the last 10% of any job that's the hardest... in this case the 1%
Who can we tag ?
thats the delima i have with giving stuff away for free... i get it but it also makes the experience worse for paid users to have to eat the consequences of all the free people generating and clogging up the server
I tryed at sonovid.ai (had some credits) it worked there.
But i want to use in flow
when free users overwhelm the servers, paying customers end up paying the price — literally
How sad the rollout is still breaking down. 🙁
please save your credits
eu
My creation
yeah paying 250 and everything is going to 99 percent aint it man lol.
quality works more often than fast though
As we wait for the magic to happen here is a song to pass the time https://www.youtube.com/watch?v=hiwgOWo7mDc&list=RDhiwgOWo7mDc&start_radio=1
Tickets: NENA LIVE 2025 https://nena.tix.to/wir-gehoeren-zusammen-live2025
#99luftballons #nenavideos #99redballoons
▶︎ https://www.nena.de
▶︎ https://shop.nena.de/
▶︎ Den NENA Youtube Kanal kannst du hier direkt abonnieren: https://nena.lnk.to/yt-sub
You can subscribe to the NENA Youtube channel here: https://www.youtube.com/c/ne...
Hi I am new i need help
german version so much better by 99% https://www.youtube.com/watch?v=Fpu5a0Bl8eY&list=RDFpu5a0Bl8eY&start_radio=1
Tickets: NENA LIVE 2025 https://nena.tix.to/wir-gehoeren-zusammen-live2025
#nenavideos #99luftballons #99redballoons #original
▶︎ https://www.nena.de
▶︎ https://shop.nena.de/
▶︎ Den NENA Youtube Kanal kannst du hier direkt abonnieren: https://nena.lnk.to/yt-sub
Generations with Veo3.1 fast are stopping at 99% again...
My credits are melting in the sun, with all these generational failures
hold on i havent checked my credits but the failed gens still charge credits?
They dont if you look at the right top corner, failed generations are immediately giving back credits they used.
Normally we have credit reimbursements but I don't have the impression that this is the case for all failures.
sometimes, it takes time to return the credits. The last one took almost 1 hour to return 💩
I wonder if any of devs are aware of that, fix it please, I cant go on with my projects!
For what work are you using it:))?
Veo 3.1 feels the same 😭
yeah im sure their are improvements but the biggest difference i notice is 99 percent unfinished videos lol
i came here to post about 99% forever. but i see its not just me
Google flow is better than Openais Sora 2?
Sora 2 looks more realistic and better physics than veo, but veo is less restricted, way faster, can do image to video of realisticlooking people. so yeah
and how much is the paid version in flow? And how many videos can I generate with the paid version.
@fading dirge
yes
Well i have the 250 dollar plan. you can do unlimited fast videos, and you have i think 24000 credits to use for quality. But right now 99 percent
250 dollar is very pricy:))
And are you making videos and posting them on tiktok, YouTube?
it is but unlimited veo 3 fast is a good deal imo. especially since veo 3 fast alot of times looks as good as veo 3 quality. all comes with youtube premium and like alot of cloud memory
I'm interested in this.
Are you making money with this?
if thats what you want to do then yeah you can do that and make money but you have to reach a point to get monetized. you wont make money day 1
Yeah, I know.
But can you monetize AI made videos?
yes you can
since they released the ultra plan
Have you published those videoes?
i have a monetized youtube and tiktok and i have used veo3 on some, so yes you can monetize AI videos
Where are you from?
Murica
Murcia?
america
aah:))
But i don't know if i could monetize it from Romania
I think I could only monetize from America
if romania youtuber can monetize, then you can too
hmm, interesting.
yeah i dont think where you are effects youteub monetization
Hello, we wanted to remind you of our server guidelines, stated in #start-here・📍 : “Currently, our team can support engagement in English; the discussion in this server will largely remain in English for now.” Please keep in mind that non-English discussions and content may be removed.
Still buggy I see. Dang. 🙁
@fading dirge @crisp basalt
There must be a problem all over the world
Any announcement about when its going to be fixed ?
Have they even acknowledged it yet?
Working again
@main saddle is the latest update of object insertion in veo 2 only ? I could check after work but thought I’d ask. Thank you 🙏
Is it already fixed?
if I utilize my Google Vids limits for 30 days then the video credits will be relevant to 6000 flow credits
so like it's 6 times more than what pro users get for flow
I just tried and got frozen on 99%. 🙁
Anyone is working on the fix? On Lucifer's sweaty balls, I can't work on my project!
I just had one that was stuck on 99% for like 15 minutes suddenly resolve and play. I don't know what is going on.
also 3 veo videos on gemini app per day is equivalent to 90 videos per month which is like 1800 flow credits. so like thats also nearly double than what pro users get. lol. I wish the limits were increased on flow
Hi guys I'm about to rejoin ultra sub, but reason I didn't continue last month was because of generating limits, I kept getting messages you are generating too fast even though I have none in queue, also is 3.1 unlimited on ultra?
Is there really hidden limits too amount of videos you can generate? Because for me it wasn't unlimited only that no credits was used.
Please some one clarify thanks.
The willow chip is eerily similar, to my "Willow" codex
Happy for the folks that can generate now, I just tried and got several stuck at 99%, and then just dropped and refunded.
I wonder if the servers are overwhelmed or they are trying to fix bugs.
Maybe they are updating the model.
it became overwhelmed since they introduced unlimited generation with veo 3 fast mode for ultra plan users and 100 monthly credit free for first time users,
This was created while I was attempting to solve p vs np
Sorry, can I know what you're referring to? Its sound cool!! I'm a pro user.
The unlimited was in place for a little while though, and is rather pricey.
I'm not surprised if they are getting hammered with new users though trying out the 3.1 model. It seemed broken on release yesterday for me. I posted some clips showing the issues I was having in flow gallery.
i was talking about flow credits being very less compared to the availability of the model in other google products
Google Vids is an AI-powered video creation and editing tool. Use it to generate Veo clips, record your screen, convert your presentations to videos, and more. Access Vids with your Workspace account or your personal Google account.
So this has happened since yesterday?
the 99%bug happened since months ago i think
it got worse now it seems
Yesterday generations started acting odd with the release of 3.1. I had double watermarking, audio issues, and found the quality had degraded (last one is a personal preference, but others agreed).
Later in the night a lot of people started saying their generations were failing.
Emmmm, I tried veo2 fast and veo3 fast 2 weeks ago, they worked well. And I just started my work again 2hrs ago and found it didn't work.
These are all the ways you can get Veo as a pro user:
- Google Vids (Veo 3.1 Fast) - 10 x 30 = 300 videos/month
- Gemini app (Veo 3.1 Fast) - 3 x 30 = 90 videos/month
- Opal (Veo 3.1 Fast) - 3 x 30 = 90 videos/month (available for all users even free users)
- Flow (Veo 3.1 Fast / Veo 3.1 Standard) - 50 / 10 videos for veo 3.1 fast / standard respectively.
There could be others as well but Im not aware of them
Yes, the quality
I noticed that 3.0 -> 3.1 the fast generations are really meh now.
I asked for rain and 3.0 did rain. 3.1 decided to put a constant rain effect on the lens. It was so stupid looking I thought it couldn't happen again, but several generations had it repeat. This is despite using the same prompt that gave great results in 3.0.
I'm not sure what's causing this though, I'm not that smart. I don't think it is just the rush of people though, I think there may be model problems as well as implementation.
Doing extend seems to take like a .5 second or so of the first clip to continue the generation, but it only seems to work well if you do a quality extend. Fast just changes the audio into whatever it would like.
I've been waiting for 10 minutes and still waiting
Why when you use the Scene builder, it's only 720p? You're still charged 100 credits for "Quality". It doesn't seem right.
Lol, brother
it always has been 720p
1080p is only when u do upscale
The only thing I’ve been getting was lack of audio on some of the videos but all of mine are generating
That's something they said was with the new 3.1. 1080p generations.
😂
I found another problem, 3.1 prefers to add music in every video.
They never did
u probably read the api veo
Just tell it not to
which the api offers option to generate in 1080p
Ah yes, you're right. I think upscaling in After Effects actually looks better. I still find the outputs from scene builder look really glossy and you have more Ai artifacts in there.
2
Got
266666
And even then the api isn’t native either
exactly haha ,
Oh, I'm sorry. I heard some folks talking about it and assumed. I only cared about extend. 🙇♂️
Topaz is goated
Facing the fierce competition from Sora2, it is necessary to lower the price, perhaps 20 points would be more appropriate.
Especially when you consider how many credits are wasted on incorrect generations
Google still allows free credits plan, so I do't think this will happen, at least this year not.
samee. after 99% its taking more time
I think the api doesn't return how much % the video is done so it still works after 99% or sometimes the generation completes early
the % I think works on an average time alloted to the videos
But when I refreshed the page, the credits refunded.
in google vids as well I got this issue. It was showing 99% for 10 minutes and then it showed me this error
is veo down? it's also not working on gemini app
While I wait, I might consider opening Sora's website. After I download Sora's 15-second video and return to flow, I can still see 99% of the gray background.
I'm also getting the same thing. Stuck at 99% for a while now
Oh my God, what I ended up doing was angrily slamming the keyboard a few times.
V3.1 should never have been released in the state it is, Google have teams of testers so what happened?, causing massive frustration to all, should be retribution to all subscribers not too mention the amount of credits that have been lost...Lucky I have backup but this is totally disgusting on Googles behalf and I for one am rethinking my subscription.
It will only make users more determined to choose Sora 2
Invite Sam Altman to this group
Can i use sora 2 with a VPN to America?
I think so
Do I need an invite code? Or something?
Yes as long as u have invite code and pro enabled, most dont as yet as its still being rolled
How much is pro?
$200
cos its new
And I can generate only 8-10 sec videos?
😂
yeah, but only for a month :))
the way ai is moving just wait 1 week and it will be $5 🤣
the sora 2 haha?
And how many videos can I generate with it a Day?
Unlimited I think? I can't pay it either.
All these companies are trying to be the first, reminds me of the Beta and VHS war (if any of u are old enuff to remember) and 1 or 2 will eventually come out on top, hence Google pushing an update that should never have been released (like the gaming World).....
I think I'm not old enough :)) only 22,and never heard of Beta and VHS
please check out my new video in #flow-videos・📹
Google it 🤣 I assume Google search works?
it seems like a lot of the updates have catered to the VFX / AI slop crowd, prioritizing crazy camera moves and improved VFX through deeper interpretation of ingredients / frames, and whisk has somewhat improved in some areas as well (though when it gets its interpretation of context wrong, it's REALLY wrong... i've been trying to revise a single CU dialogue scene all morning that worked perfectly fine with Veo 3. Today, it's turning my 1950's retro homemaker character into a pageant contestant on a reality TV show, complete with sad music, dramatic camera zooms, and a modern gen z accent that won't go away.
These bugs (like 99% stuck) are killing me though. I have a project deadline tonight and if i hadn't gotten a head start on it, i'd be screwed.
In immediate rollout, I was really enjoying the improvement to ingredients and frames, and it's good at making things look "cool" but that's not what serious filmmakers / businesses need, especially long term.
Is there no support here?
Yeah, the best approach would be leaving the option to choose whether to use Veo 3 or 3.1. I mean they still keep Veo 2 for a reason right? Huge mistake on their behalf. ^^"
Doing this after Sora2 makes it feel like a rushed response.
Having it be so buggy and breaking on release while Sora is adding new features to get parity makes it feel panicked.
Why can't fast be used for jump?
Exactly this
Hello! Our apologies for the rocky past couple of days. Our main issue here is capacity (available resources vs user demand), rather than anything that's specific to the Veo 3.1 model rather than Veo 3.0 model. FWIW this wasn't really timed around Sora, releasing this kind of thing at Google takes a long time and it's been in progress since before that announcement.
Interesting, I thought that should work, https://support.google.com/labs/answer/16352836?hl=en&ref_topic=16355775&sjid=12641989696427378568-NA claims it will, let me file a bug with the team and see if we have an error somewhere
Flow provides a range of features powered by various Veo models, in addition to Imagen 4 and Gemini. Different models support different features. Veo 2 is our default model and supports the full F
thank you
Another bug, youre be running out of bug server space
Kind of you to reply. I totally understand the rush of people causing server issues.
Yesterday when I got to try the videos I noticed though that the extend feature caused a lot of issues and redoing generations with 3.1 Fast seemed to have worse outputs.
Are the bugs some folks like me are seeing tied to a rocky rollout, our bad prompting, or is this just how the new model looks?
I posted some clips to demonstrate what I saw yesterday in flow gallery.
Hi, my opinion is that the generated video should be without soundtrack by default.
If you're seeing generations not completing at all or stalling at 99%, that's the capacity issues. If you're seeing that you're less happy with the outputs of a model, that's tied to what Veo 3.1 outputs compared to Veo 3.0 and we'd be very interested to get more information there. Extend, particularly audio (which we will label beta) has been one we've gotten more reports on recently, I'll look at yours in the gallery
ugc
Thank you very much. I can post more details if needed including the prompt used and more samples.
I'm super eager to get some longer video ideas out once I can. 😄
Good suggestion, we've considered making that an option, but it's technically quite hard/annoying for us to do, so we haven't prioritized it, and we assume that users can (if they really need to) go and remove the audio track elsewhere. Not ideal, just us working through a stack rank of priorities.
So when will the capacitry issue be fixed since everyone is complaining
When the new people get tired of failed generations so the rest of us can get them 😆
@thick oriole Is it possible to reduce the price of Quality?
Wait but why it's competing if Veo 3 lets you choose scenes and manage everything while Sora 2 (if i'm not wrong) has only space for a prompt with no control options? Could you share some information about it?
I don't have any timeline. Many many folks are urgently working on it. A guess (not a promise, please don't take this as anything official) is by later today or early tomorrow.
Okay thank you
I'm gonna guess that is a decision that's many levels higher than this discord
The competition is too fierce, so it is inevitable to bring benefits to users
haha, I can always try, but I'm more on the engineering side! My vague understanding is that generally, when we're already over-capacity, increasing demand by reducing cost isn't very desirable. But equally, we'd love to reduce cost in general (often by making the models cost less to serve) to attract more users, more usage, etc.
I believe Sora 2 will put pressure on Veo3.1, so we will soon see 20 points of Quality
This is honestly very similar to how almost every AI update release seems to go. Everyone rushes to see the new features and end up demolishing the servers.
There are a lot of neat things Sora is doing, but they already said that the current rate of generations is temporary and they'll be limiting them soon.
Their idea of rolling it out slowly through invite codes was a good call to prevent their servers from melting down.
+1
This is the way I think Google will do and make it a better Google.
yes but releasing a product not fit for use can do more harm than good, i.e you only have to look at what happened to the game "no mans sky" on release (to a mush lesser scale but same principle), luckily they made it up on all future update's that have been FREE, but at release it tanked...so Google or no Google they could lose this battle if they aint careful as there are a lot of big players...and I cant see GG giving anything for free...
is flow currentlynot working?
yeah
a refresh loaded the generated 99% videos for me 😄
Server capacity issue. They're trying to fix it.
i have sora 2.. but it was watermark
how do you even have sora 2
Where are you from?
Top 3 in the World for server cap, just a joke...
I'd assume from an invite code. I've got it and it is "OK."
I'd say similar in quality to Veo3 honestly. 🤷♂️
Jump To not working with Veo 3.1 - Fast
Its all down to how you write and organise ur prompts, Veo completly different to S2
You say that Veo is better?
Haha same, hope this get fixed
What project do you have? :)))
A fresh refunded all my credits cost on 99% loading videos.😂
Veo is for film making it’ll always win
Actually just for a youtube channel of my client and also my own channel.
this works lol, but hope they fix this as soon as possible.
... it depends on what you are doing honestly.
I've got some prompts that in Veo come out meh, and in Sora come out great. I've had the reverse happen as well. It seems like Veo responds best to longer prompts that describe what happpens visually. Sora does better with more technical ones. Like a prompt written in markdown format and very code based.
I have to use an AI to help me write in prompts like that though since it is such a pain.
ok rant over by by Veo, hello openart lets get some work done
Until 3.1 starts working again I can't really compare and contrast though.
This content may violate our guardrails - about half my generations there anymore 😆
flow is still broken??
on my side, yes.
what am i paying for at this point 💢
its stuck on 99% :(
They'll probably fix this soon, I just hope they fix it as soon as possible since a lot of people are affected.
well it broke at like 2 am last night. still broken after ive slept, and im just most shocked that none of the staff have even acknowledged the errors
oh wait it worked for me it just took god knows how long on 99%
There's a thread about it.
okay i did get a result this time, that happened only 1 time last night
They are and have. I hope they put something up on the site so people feel like they are working on it. (Even if they can't do anything, it'll make people feel better)
Anyway I dont care anymore, cos ive been upgraded from a Little Snail to a Lively Otter, job done.
okay, i didnt see the thread
ty
It works perfect now, on my account.
omg, i just realized we have ranks 😆
Its my only reason for living
I outrank you, so step in line son
Aye aye 🫡
You tricked me into trying again, you rascal you
-# edit: it didn't work fyi
yeah V3 is working again woz tryin to keep the server space down
the soundtrack thing has ruined a lot of clips for me and wasted a lot of money / credits, and based on the little I experimented with last night / this morning with Veo 3.1 - it seems to be much worse now. I'm strongly considering switching (ending subscriptio) if this isn't going to be fixed any time soon
Same.
No dice for me. I tried on gemini and flow and they failed. 🎻 🥀
Go for a sleep guys
Ugh, guess I'll go and do work then 😭
its working 1/3 times for me
Hey, be careful testing your generations right now. Maybe look at your credits before and after. I just realized my failed gens haven't been refunded. Only lost 20 credits or so, but 😨
i've been monitoring it DW. i haven't lost any
i was super worried about that earlier too though
3 failed gens in a row now though, maybe it broke again :/
It has only happened to me once, my other failed gens got refunded. I'm gonna assume glitch that will be fixed.
EDIT: I've been refunded, it just took six to eight minutes. Server is crashing 🔥 🧯
hope so! that's not cool
i'd feel bad for those who were trying quality model. they might be out 200
🫶
is %99 problem still going?
does anyone else experience changes in nano banana? I'm not able to perform simple face swap request anymore...?
See the announcement
i don't see it tldr me
Google is experiencing capacity issues across many GenAI models due to the large spike in demand with the launch of Veo 3.1. We're urgently working on fixing that, and we know it's impacting users of Flow and Whisk. We don't have a specific timeline to share.
thanks
its ok google
Can’t wait to try out 3.1
yea i mean the issues should fix themselves after a day or so once the hype dies down and the normies move on
Man, I'm itching to get the extend feature to work. To be able to get clips to the 20 second mark before a cut would be sooooo nice
Right now anything I make has an ADHD feel to it with everyone talking as fast as possible.
yeah im having the same experience. it feels so rushed
i would gladly give like 30 tokens/gen for a longer clip
Flow announcements will always ping the Flow role. feel free to grab this role under "Channels & Roles" if interested. for now, here's the latest: #flow-announcements・📣 message
Please im new here how can i do to use images and videos in flow
ugh every video is so hyperactive. i cant get it to slow down or look normal
switch it to frames to video or ingredients to video
Thank U so much
Why is the gemini image editor quality so bad
bruh the 99 percent thing still aint fixed? lol
hi! any teachers out there? curious how educators might be using Flow for learning in the classroom or elsewhere 🍏 ✏️
Working for me now, is it still an issue for you guys?
I'm just burning through credits making videos with no sound. When it works it's great, but when it doesn't it's useless.
Is there any official news what exactly changed or improved with veo 3.1 ?
Hello, we wanted to remind you of our server guidelines, stated in #start-here・📍 : “Currently, our team can support engagement in English; the discussion in this server will largely remain in English for now.” Please keep in mind that non-English discussions and content may be removed. Thank you!
check out yesterday's [Veo 3.1 announcements!](#flow-announcements・📣 message)
hello. Is it possible to remove Veo watermark from a video?
hey there! videos generated under Google One Ultra plans will not contain a visible watermark. for more info on watermarks, check this article out
Create cinematic clips, scenes, and stories with consistency using Flow, our AI filmmaking tool. Flow was built with and for creatives and uses Google’s most capable generative models.
hey Tobie - [we hear ya](#flow-announcements・📣 message). we're working on resolving these issues as soon as we can!
hey Georgia, sure no problem 👍
thanks for all your effort
flow is back ?
yes
Google flow you charge me for flow. On the 14 but I cancel it before the 16 when it was due. But it seem you took my veo 3 flow away. Question. Will I get my money back or will you help me figure out why I was charge that. Thank for your help. If anyone have any ideas let me know what to do to solve this issue.
hey there! thanks for reaching out. I'd recommend contacting Google One Support for help with this as we're unable to look at your account thru Discord.
Thanks
I've had an issue, where if i create 2 videos per prompt 2nd video usually is completely random, character says random stuff, or looks completely different, while the first video is normal
Any suggestions as why this might happend?
Hi im new here
Any recommendations? Someone to follow or any IG/YouTube channel to learn Google Flow?
When generating videos with the new model VEO 3.1 no audio is generated, do you know when this could be solved?
what next week ? wow.. and how about ultra plan ? a week its too long
Please someone help me with this, every time I use the extend feature it always says something went wrong, this has been going on for every single video generation across veo 3.1 extend to. Please someone help me with this.
With support ending soon for Veo 2 (fast) and Veo 2 (quality).... will there be a cheaper option still for video without sound? Or will all videos cost 100 credits??
The Veo 3.1 Fast cost 20 credits per video the only expensive one is Veo 3.1 quality for 100 credits.
Also does anyone know the right prompt to make the script the AI should say? I keep making one but it keep changing. 😭
With 3.1, I feel there's been a downgrade with dialogue, it has a mouth of its own now
but since the beginning, this has worked for me pretty good --
John (right) asks: "What restaurant?" Lisa (left) replies: "Cheesecake."
I know that may look simple...
Audio for me, since the update..... is almost (1) I get no dialogue audio or (2) dialogue with background music
Even the camera has a mind of its own
anyone have good results with stuff like "static camera" or another phrase??
Thanks I just make a video with a script provided and it just put some shopping mall music on 💀 I'll try yours ^^
Yeah it's so hard to get voices now
Okay I just tried stuff and the only thing that help me with the script voice is asking Gemini to give the prompt
[AUDIO]
Voiceover: A confident and energetic male voice in English says, "Check it out right now, it's freshly made."
Sound: Upbeat, modern, and royalty-free background music.
nice
My guess is they are tweaking how extend works since it was sorta buggy yesterday.
What are the chances that I'll be able to generate a sweet and innocent kiss between the main characters, when I eventually get to this part of the short film?
I'm doomed, ain't I? Using this Google product...
I've noticed extend working a bit better today, so that's nice.
I hit a brick wall today, I think 3.1 is broken
It is... a rough start
So, I'm sure this has been talked about, but is there a reason that Veo doesn't really allow for famous figures in history?
Tried doing some videos of Nixon or Kennedy and... they are not them.
for me the dialog, sfx audio all good. but just negative prompt
Veo 3.1 have been working wonders for me idk
Same. Much more realistic and detailed for my use
is flow working now for everyone?
It lags pretty badly for me when I use it on chrome…but Safari is flawless.. oddly lol
Do we still have music in every video?
99% 99% 99%
do quotas reset every 24h? had to swap from flow to aistudio
guys i need help
i was in the storyboard mode but i cant get back into that mode
i click add
how do i go back in to the extend thimg
@cedar bloom
pls
help me
@sly sail help pls
Yes I am very impressed
flow is the thing thats broken. the AIstudio page for veo 3.1 works flawlessly for me, but flow is the only thing that's breaking/getting stuck at 99%
help me pls
have you considered the extend button?
idk dude im not a staff member
is that ia good for anime scene??
wanna make gojo vs sukuna
you aren't supposed to be making videos with copyrighted content
make you vs me
in anime
please stop pinging me
OK
Thank you flow it seem it was handle you reversal the payment. I kind of wish I could have kept it. For oct. I don’t want to loose the discount. I hope I can still do it for the two months once I get the payment again. Honesty I think flow should stay around $100 not $200 but that up to you. You’re one of the best out there. (Tho you do have competition but that a good thing not a bad thing. I will remember that you must cancel your subscription 5 days ahead next time I had no idea you would charge me on the 14 of October tho the renew date was on the 16th thanks again flow for helping me and google team.
hello 🙂
can someoen explain to me how to use veo3 ? All I have now is a pro gemini account so I'm locked behind 3 videos a day
And this entire pricing structure is so complicated that I'm lost.
Please, can someone explain to me like I'm 5 ?
I need an unlimited plan for veo 3 for professional purposes.
anyone else getting this
what part of it are you having trouble with
- I heard people say we can have a free trial with 300$ worth of credit
- apparentl ÚI need ultra plan to access veo 3, but I don't see how much video I'll be able to make
- If I go beyond the limit of ai ultra, how much will each video cost ?
- is ultra the only way to access veo 3 ?
- whats the difference between using veo3 in flow vs gemini
- is there another "place" to use veo ?
- is there a question I'm eventually missing that is worth asking ? 🙂
you'll want to use https://ai.google.dev/gemini-api/docs/rate-limits to view how many videos you can make per your plan, and the pricing per type of video generation here: https://ai.google.dev/gemini-api/docs/pricing
there are 2 places to use VEO afaik.
One is flow: https://labs.google/flow/about
another is in the ai studio app: https://aistudio.google.com/apps/bundled/veo_studio
AI studio app uses the $300 credit you're talking about ($0.15 per 3.1 fast, $0.40 per 3.1 normal), while Flow uses different credits (20 per 3.1 Fast, 100 per 3.1 quality)
so you'll want to be using the ai studio app if you've activated the free trial through the cloud console
apologies if there's a slight amount of misinformation, im not a staff member or anything, this is just what i learned from using veo
okay im sorry but im getting kinda pissed. i get 20 of these per day and all 6 requests were wasted because of errors and were not refunded like the Flow system.
anyone finding 3.1 not working?
here we go again
thank you very much for your time its clearer now !
what would be your plan if you wanted to make around 500 videos a month personally ?
with how inconsistent it's been? i dont know if I would spend that much time and money on this until they work out the issues
it's broken all the time
tungtung sahur
see?
lol
good timing lol
its the best tool out there for me tho, I wanna make 3d animation and ysterday it did an impressive job
I tried sora and other tools, its absolute garbage compared to that
yeah it is really good when it works
perhaps I should wait 1 month for it to stabilize
I'm not in a huge rush anywa<y
and perhaps prices will go down over time too
dont expect that mate
these AI companies will not lower their prices and keep on delivering broken updates and not have them stable until like 1-2 weeks after
to be fair if a new veo model releases the credit/gen cost might decrease for 3.1
happened with veo 2
Is anyone else getting the "this video model isnt avalible at this time" message, having issues with their projects displaying, and or generating videos at all on VEO?
yes it's likely traffic overload again
i can generate animations just fine right now btw, but I use the Fast version not the Quality
can we also count on competition to lower the price ? Sora 2 is super good (althought not for my 3d use cases but yea)
It's broken again. I see why OpenAI geo-limited access now.
and i assume it has a little something to do with this. it's always breaking near midnight PST.
everyone starts generating at midnight PST. servers get f'ed
it's 9am in EU, na people should be asleep now but I guess eu timezone is the highest traffic time
Whats the ETA yall think they'll fix it?
you make 3d animations ?
I don't think they fix it at all. It's just when the traffic dies down again.
yes
thats probably why
America is sleeping so traffic is prolly lower
oh so never then? oof
no we arent
traffic is not low, it's extremely high. that's why it's broken
can I see what you do perhaps ? I'd love to learn how to optimize prompting with veo as I'm new to it
i used it for past 3 hrs successfully now its failing
Just my guess. They don't maintain a Status thread here or post immediately as outages occur. Maybe no real focus on Discord.
Can't really show what we do, but I'm learning too mate we just swapped over to Veo from other AI animation programs so I'm no expert at it yet, figuring stuff out everyday
What are you working on?
Oh no is it down again?
makes sense yea, basically making simple animations from image to video for my video game studio. The background is empty and I need to animate it so that the character stays in frame, and if possible make a loop
Making it stay in frame is the hardest challenge to me, and having a natural movement
I don't know if we can change the aspect ratio in veo 3 aswell to make it easier
well if my character wants to raise his hands, move his sword, stuff like that yea, the sword will leave the frame
right, my png is 9 16 so I need to manually change it to 16 9 before uploading it I guess
Hmm, and you need it to be a full body shot throughout the whole animation I assume?
Have you tried with the image as is and then just choosing 16:9 in VEO? I've never tried that as our images are already 16:9, but it might work
Is flow broken again? God I'm already feeling my blood boil 🥰
we might as well go camping for a week
A lot of people who are paying subscribers is paying for a service and if your not providing the service it's definitely a middle finger to the customer.
ESPECIALLY for people who use the service for work.
yeah well OpenAI dropped Sora 2 so they had to rush out 3.1 I assume
always like this, one company drops an update and others follow suit shortly after and the update is brokeeeeeen
its pathetic really
So because the server is full is why i'm getting the "Video model unavalibe at this time message" ?
Hello guys 👋
flow is still broken I guess 🥲
Some acknowledgement would be great, especially since yesterday it's broken for at least 6 hours and today it's starting again.
Good thing I'm a pro user, imagine the ultra users and STILL having the issue.
I think the server is just under heavy load, when something new comes out, everyone floods it. Im an ultra user and yeah im mad lol but i understand why this is happening
i cant even tell you if 3.1 is better because i cant generater videos
When times like this happen, i get a little salty that people get to use for free lol because over all that makes things worse for the paying customers
Yeah, I can't help but resolve to alternative like Klingai because at least they prioritise the paying subscriber where free user don't get free stuff much. I would rather pay for klingai subscription and just keep the pro for Flow if this keeps up.
And they would not do this kind of situation where they just not announce if there's bugs or error with the server. Truly unprofessional, I don't want to be a Karen but I'm just tired trying to work with it. ☺️ 🙏
Yeah and honestly outside of sound, Kling does better than veo, but man having good sound is a great thing
@fading dirge what do you do? Asking because you think Kling is good. We swapped from Kling to VEO because of mainly lip sync, but also because Kling wasn't very good and had a lot of morphing
i dont blame you man. with how expensive google's services are, paying for a premium should result in premium services. that means... not this
if lip sync is what toure looking for, veo is much better than kling for that
its just alot of things kling can do that veo just cant as far as animations
tbf we haven't used Klings 2.5 version, how is it?
alot less warping, kling 2.1 started good but then its just became bad, like warping way to much, thats mostly fixed in 2.5
alright, that's nice, that was like the biggest issue, and the lip sync of course
If Kling could fix perfect lip sync like Veo then it could be viable for animated movies
yeah, its sound and lip sync just isnt nowhere near as good on kling as veo. But yeah im payinh 250 a month and credits dont carry over on veo so its like bruh, for this to be happening has me salty lol.
yeah its not ideal is it
Is Veo still bugged? It stops again on 99%
sounds like it reading above
i can get text to video to work but not frames or ingredients to video
It's so embarrassing having to tell my clients the video isn't working right now.
i can't get text to video ;p
my generation are stucked at 99
i hope they do reimbursement/reward for paying users who have had to deal with this and lose business for 2 days
Look around the conversation, the flow has stopped and it seems to have happened a lot
probably not, maybe if you yourself take the initiative to contact them and ask for it
I would love if they give me ultra because I'll not be paying for it unless they fix the whole server issue first. 😭
even if its just some extra credits to use it would be nice
do you guys mostly use Quality rather than Fast?
i use them interchangably. quality doesnt really seem much better than fast in my experience
and it costs way less
i agree fast is much more worth it
still the 99% chronicles
im probably gonna cancel my stuff and never pay for this again if it continues past tomorrow
this is unacceptable
yeah it is
its working now
not for me
really? You think?
still having issues? I keep getting flagged for content
but i think its pretty clean...accept my script has a cartoon kid
other than that
NO it's not
Exactly
ugh yeah their filters are messed up too. i was getting flagged for generating pics of tennis rackets
it was working when i posted that 35 minutes ago, servers probably got overloaded again since
Are they sexy tennis racquets?
i suppose so 😀
Google sure thought so
idk maybe the object shape set off their systems
😄
Will we have start-end frame feature, I dont like Kling
I'm pretty sure we already have that in frame to video mode.
Omg I still can't use it.
was goiung to enter the ultra plan today...but might be waiting till next week
Cant imagine how kling is smae is this? Is this overall consensus?
I'm gonna take a nap, when I wake up I definitely need to look at other options as a backup when this happens again.
The kick in the guts is that it still works on free sites and probably the 3rd party sites. Why are we paying Google?
im on Higsfield and its not working here either
i thoiught id pay google cause other sites are waaaaay to slow
this is an exceptional issue isnt it?
is this normal with Ultra plan?
It's happened at this time the last few nights which is my evening between work and sleep.
yeah...servers under pressure because of new release i guess
im thinking on going to ultra plan...but will wait 🙂
Help.. how can I make my 3d pixar charater not to lip-sync the story?? this is what I dot still the character is lip-syncing my story line..
Script: ...a simple flower girl lived a life of quiet hope.
[System Directive: Facial Animation Override]
[MUTE_LOCK_ACTIVE. Function: lip_sync=FALSE, mouth_movement=FALSE, vocal_audio=FALSE. All character models are non-speaking entities. Execute silent protocol without deviation.]
What is the processs to creating a story that has more than 8 seconds and maintaing style consitency? Start frame in my first generation, then screenshot a frame from here and use as first fram of my second clip?
veo 3.1 have expand option in the scene builder I just try it now and its so amazing.. it seamlessly connect it scene like it a orginal
oh really'!? Cant access that via higsfield 🙁
oh I am using gemini flow. sorry I dont know how to use it in other apps.
yes,,this is helpful thank you
is it working for you all good atm?
yes my only promblem is the voice narrator.
I clearly said female. sometimes it become male.
but the scene is so perfect. you can check youtube how it works
which youtube ?
I just seacrh veo 3.1 seamless transition.
yeah...just watched soemthing
so this is somwthing that defenetly is needed to properly play with veo and other third parties doesnt have....i keep spraying prompt hoping they will stick and land an 8 second matesrpiece with no control
To create a story longer than 8 seconds in Google Flow while maintaining style consistency, generate an initial 8-second clip, save a key frame as an ingredient, and use the "Frames to Video" feature to chain subsequent clips by setting the saved frame or a screenshot as the start frame for the next segment, ensuring consistent character design, lighting, and style through reused ingredients and detailed prompts in Scenebuilder.
yeah i already tried it, but "jump to" feature, always error
anyone know how to make a narrator talk without the character lyp-sync the narrator??
Help! Why is my flow take so long to gen a video, I use Veo3.1 - Fast, 9:16, 1 output
i would defenetly skip that part and narrate with elven labs if you dont want lipsync
will give you lees issues im sure
you just need more prompt @hexed socket , do prompt engineering. for me veo 3.1 dont have negative prompt. but when you can do better json prompt, you will get all audio n sfx sync in 3.1.
look the last one i share on flow gallery
Hi everyone,
I just wanted to confirm whether it’s allowed if I personally remove the Veo logo from my videos if I have a paid Google Pro subscription, the videos would only be used for personal purposes, such as posting on TikTok or YouTube?
theres logo on the videos if using Ultra plan?