#general・🧺
1 messages · Page 5 of 1
Yep waited long asf but wont get probaby till next year and will be open acess to everyone and the waitling list wont matter
hey when does labs be availabe for india
Hmmm
I have faced the same problem. Is there any way to use it from Bangladesh
video/
yea i bet its gonna be neutered and a low monthly limit too :/
Hiya, I am new here. How do I report a serious bug on the Google labs website? I fear there are visual bugs which could induce seizures to people with epilepsy.
Thanks!! Is there also a more urgent way to report it? Because I fear it has potential to be dangerous.
sadly that’s the best you could do to report it
or you can dm it to one of the team members
https://discord.com/channels/1172568727942860810/1320232517919440927
Its apparently already been reported. But is not yet fixed. Reported on the 22nd of December.
the head of community is @raw bison and the head of google labs is @strong thunder
not sure if their dm’s are enabled
you could check tho
prolly contact jamie as he might have access to the website
Thanks! I DMd both of them.
What’s wrong with this word?, I sense racism
It’s out of context
But why would that word be blocked?
What if you want to use it in your prompt?
Like white is not even blocked
I signed in on labs.google with my id , and accepted i would like to get email, does that inlisted me on waitlist to use veo 2 ,if now please someone guide me , i really wanna use it..
Sadly you do have to wait it out
The time that you get access varies between people
some get it early , some get it late
But i have heard rumours that veo 2 is releasing early january of 2025
Can you show the source?
Where the rumours are from?
Idk i just heard it from @swift jolt
It’s only allegedly
Ok man , i just wanna know if i need to fill a seperate form or something for waitlist,or i am just listed on waitlist by signing in ?
Seperate form to join waitlist
Where can i get it ?
@scenic turtle why do you have onlyfans in your bio bro
Hehe try joining it and you will know 😜🤣
Jeez I did, it scared tf out of me
classic screamer
I’ve seen that link already
Ok thanks for this, I just remembered i filled the form like a week ago😅 , but i will fill it again, i guess that increases chances
Forget it till next year 😂
what do i do with 470 120fps gaming clips?
38 gigabytes, I am just uploading all to youtube but pretty sure I lose some quality and fps
Hey, anyone know if I can use ImageFX images for commercial stuff?
Creating a large variety of high-quality image assets can be challenging, so Google Ads offers generative AI tools to help you create new image assets in seconds. You can partner with Google AI to br
Tks! I´ll take a look
Hey this is a trailer i made using videofc: https://youtu.be/8Cbow0xehxQ?si=PrHE9g8tMh3GTiRw
( Google's DeepMind Veo 2 Generated Trailer)
When is the whisk available in Australia or elsewhere?
if your using vertex API you dont get charged for blocked generations, if you are using labs its free so there are no credits to refund 😎
Lmfao, I was not expecting the link to do what it did but I love it
Now make it a virus 😈 (joking do not)

love the detalization on imagefx...wondering what model used there?
Imagen 3
I reeeaally wish I had access to the new video generator tool. It looks like it's good and dancing
video requests are usually done in #flow-videos・📹
@subtle vessel is one of them
feel free to ping him if you want your requests done
can I? thanks!
ping me in #flow-videos・📹
Just wanted to say. THANK YOU Google 😭

how can i remove a prompt "block"?
press the left or right arrow key until the line is next to the word and then press backspace
i cant for some reason( need to develop that thing
hmm
Like click the text box as normal
And then press the arrow keys
it just sliding through my generations)
an X button would be great solution) i think
idk
prolly wait for someone else to answer
thnx for the reply
prob ill wait for the update
np
Will the video ai become accessible to all?
I heard rumours from someone that it releases early January of 2025
For free? 🥺
Good day Google world
Will V2 be available globally?
soon i think
Lovely
I don't like how imagefx forces every image to be a certain quality style, it doesn't let prompts like "noisy" "amateur camera" "barely any light" have any effect on the image
it can be very hard to achieve certain art styles due to this
It depends on your prompt. I can achieve this

Where
Show us
might have to play around with the settings then or use a different term that pertains grainy / or handheld camera look
question: why sometimes I can edit the generated images and sometimes I can't?????
......no, really. How do I enable the image editing (inpainting) again??
._. no one knows?
generate these images
@subtle vessel added some proper music to the dance, it looks and sounds great!
so...
why is the inpainting not working?
I kinda need it, it was very useful
"amazing cartoon tv show screenshot of a female humanoid parrot, teal feathers, red crest , beak with teeth , many large feathers on her winglike arms , clawed hands, wearing pirate clothes , looking confused , as she is with various different ugly human pirates that are bigger"
this prompt gets unvariably blocked :/
oh I hate stupid filters
let's try to make her "happy"
...nope. unacceptable prompt
also "talking with" doesn't work 😭
please fix the filters 🥺
wow. even if she is alone it doesn't work???
"amazing cartoon tv show screenshot of a female humanoid parrot, teal feathers, red crest , beak with teeth , many large feathers on her winglike arms , clawed hands , wearing pirate clothes" -> flagged as nsfw
it's really odd that I can do horror movie scenes but a cartoon of a humanoid bird talking with pirates is constantly blocked....
It blocks almost everything with female characters
You can use that same prompt as male character easily
Also I never heard of in-painting imagefx, I think that is in midjourney?
or a feature that’s not added yet? Or it could be a pc thing because I’ve only been generating on my phone
lol, even a bird. an actual bird. just because it's female
solving potential sexism by making more sexism, maybe
Yup I was trying to make a Christmas card of a female artist holding a cat and I couldn’t even change eye colour or hair colour
now this is something that needs to be fixed
pretty disheartening that the human nature is to find women offensive
I wanted to make a Christmas card for my friends
I gave up because wouldn’t let me change character to have the right hair colour
Apparently changing the light green to dark green it’s filtered
Now it’s ai learning to become like humans 
AI has always mimicked humans and society, I guess it's why people panic and over-censor it instead of, y'know, focusing on the society
how can I create image with whisk
genuinely think i generated over 10k images with imagen
🤣
the wait for veo is painful
is the waitlist being accepted yet?
Is there a video tutorial available for learning how to deploy and use the various AI agents offered by Google?
Guys I want to trasform a image of myself in anime style but i dont know which program to use. What you suggest to have best results ?
Whisk because u can choose a subject from a picture and then change the style to anime
And u can also choose the background from an image
vertex, model garden.
The API documentation should be enough
👆
I wasnt sure what they were after, since ai agent can mean two things lol.
well three, for those that have zero dev exp lol "ai agent" = llm
pretty sure my irl friends intentionally say the wrong stuff to drive me nuts lol
Me too, but i guess the info we provided should be enough
happy to give you a crash course (brief-entry level) on a topic you'd like to discuss. dms are open.
sorry xD im getting all confused.
Cane someone tell me why "real-life footage of a real accurate robin bird , wearing a christmas hat . she is standing on a nice wooden table , next to a big panettone cake and a champaigne bottle. there is also a glass full of champaigne."
....is not safe for work????
oh maybe it's the "she"?
omg. yes. it was.
I'm... speechless
No. It actually still thinks it's unsafe. My bad
it was the word "panettone" 😭
I have noticed an issue with ImageFX: it will never do real versions of things that aren't real
as you can see, it won't do a REAL version of the monster, but always a drawing, no matter what I specify
but there's more
crows made of darkness don't exist, so it NEVER makes a real crow but always a drawing
It keeps happening. I can't make fictional creatures unless they're in a non-realistic style
I find this to be a big issue personally
compare with Ideogram AI:
Ideogram AI can make real-looking fictional creatures without problems
It looks way more real than the one made by Imagen however
Avoid using "real photo"
Instead, try using the camera lens or info about the camera
Mmh I will try
Yes it does make real version of fake things
Like 55mm photograph of a....
Got it
@topaz zealot I made some tests
by asking for a photo it invariably makes something that is real
so, no monsters
I can't make a man whose body is made of clockwork machinery - it always makes a man with clockwork tattoos
it's not very creative, is it
you ask for something lifelike - it's perfect. you ask for something different - it won't do it
"movie scene" makes it a bit better because it knows movies can have unreal things
I should make some Ideogram vs Dalle3 vs Imagen3 comparisons
sure
gives an insight on what certain ai models are good at
Depends on what style u are asking for and also how u describe the prompt
It’s good at making half cat half snake monster
hello, is it possible to change the image output format on Whisk please?
as in like png and jpeg?
not that i’m aware of no
Happy new year besties
Love-Truth-Compassion
does photorealistic work?
or something along the lines of "high quality movie screengrab of a photorealistic scene where [...]"
i think so
if you include 35mm film it’ll look like a movie
here’s an example
fascinating
i didnt know that!
Mmmh I will try
try what htmlll said too! 35mm + what i said
maybe it'll give more interesting results!
ill test it myself too
"High quality, beautiful, and 35mm film movie screengrab of a photorealistic scene where there's a close-up of of a photorealistic crow made out of what appears to the pure element of darkness itself, the scene's mood is intense."
i may have been a bit on the nose on the prompt but hey, it works :p
still not made of darkness, but it's beautiful
thing is, in the AI's interpretation, it is made out of darkness, just not with the effects you wanted
for it to have the effect you want, you'll have to specify said effect
it'd try it but... genuinely don't know how to put that into a prompt. words are not coming to mind rn
you can use chatgpt to enhance your prompt
or deepseek
Good point
yeah but how can i do that if i dont even know the words to use 😭
ill just screenshot the crow in the og picture and ask for a description, then put it into a prompt
Maybe.... "the crow appears to be a shadow"... idk
are people getting accepted for trusted tester yet
i think someone said late december
now its 12/31
why is videofx telling me to sign in when im already signed in?
that seems to have done it
"High quality, beautiful, and 35mm film movie screengrab of a photorealistic scene where there's a close-up of of a photorealistic crow made out of what appears to the pure element of darkness itself, manifesting as a shadowy crow, the scene's mood is intense."
It's perfect!!
its all about prompting at the end of the day !!
when making a prompt to imagen3, remember that you gotta consider that it'll take it in the most literal way known to mankind, and it won't do anything beyond what you ask it to
sometimes it does, but when it doesn't, you gotta specify what you want
real
tbh that's smart
think they said they're aiming for public access Q1 2025
Ok
Thanks for the tip!
I wonder if there's anything official mentioning this, even a msg in here from staff, I believe u and it's what I've heard but not sure where I should set my expectations lol
It is true that Google will release access in January ?
happy new year
Happy new year
happy new year:D
it’s only a rumour
Yes really, I created it now
ok
This ai is like peak laziness and tbh I relate to it. It’s like I generated this image once before for some guy, and u want me to make a similar thing again, here have the same image 
Happy new year to you all 🥳
Happy New Year!
Happy new year Google team and everyone!@
lol everyone lives in EST
Happy 2025!! :D
happy new year everyone! 
yeah
I think u can generate 400 images a day
ah, I actually made 400??? wow then my bad
400 is a good number, then I won't complain
Happy new year!
hah no
its lower
i did 20ish generations today
considering its 4 images per generation, it was about 80ish images
nah its not 400
It’s not 20 generations, it’s way more than that
You probably just already used them yesterday and the day hasn’t reset because it’s probably a different Timezone to yours
yesterday i did around 15-22 generations
that would make it 42ish generations
its still not close to 400 images
I saved about 230 images and they were just the ones I liked and that worked, there were more that didn’t yesterday
what do you need help with?
it does this every time i cant use it 😦
and
it gets really rude
oh uh
i have tried but results are always weird
try changing the system instructions?
its empty
see look here its empty :((
i just want to test it raw to see how good it is but i can even
gemini?
Oh wait your using the thinking version
yeah
i cant do anything :((
I’m gonna try replicate it
that’s weird, doesn’t happen on my end
Wait lemme try a different question
ill show with a new chat
brb
It did the prompt perfectly for me
wait there
ok
its workig now
what was the issue?
oh ok
ty for your help tho
but i really do know what was going on there?
or you dont know?
nope
ok
you know what i mean?
yes
what?
trolling
Is there any place where we can try Genie 2?
no. its not for public yet
Genie is a text to 3D model by Luma Labs
I can’t wait to play video game about arctic fox in space
❤️
Using proper spelling, grammar and sentence structure is a crucial part in the creation of prompts.
turn 1206 up to 2 and hold onto something.
problem was solved already
i realize that, i was merely giving feedback about the funs of changing temps.
ah ok
1206 @ temp 2 is legit... scary lol
I accidentally wrote blackbeard instead of blackbird
it's true lol
it's my fav model to create personas on tbh
it's too overcensored
I can't even have a medieval peasant boy talk to a female knight.
Google pls...
makes me want to return to Ideogram, which is censored as well, but not to the point of blocking ANYTHING
"movie scene of a plain-looking 40 years old medieval woman with brown hair, wearing a heavy real medieval armour that has a stylized araldic yellow eagle painted on it. she is talking to a teen with unkempt blond-brown hair who wears a simple burlap shirt"
this is blocked by the filter
about to lose my patience, it's a bit like Dalle: only good to make animals
Don't want to see image fx in Trusted Testers phase. It was very very censured.
I'm envious
Trusted Tester
but you tell me it was..... WORSE?
Yea. Coudn't even generate a cat playing with ball of titanium. But i sended the feedback to the labs and they solved the wierd overcensured preety fast
There is button on the top right of the screen to send feedback
Glabs is so cool!
Hey everyone! I'm really keen to try out Veo2 and see what it can do. Does anyone have a test account they'd be willing to let me borrow for a bit? I'd be happy to pay you $20 for the trouble. Thanks a bunch!
video requests are usually done for free in #flow-videos・📹
@subtle vessel is one of them
you’ll know when it happens tho
Does Google have an equivalent for ai coding helpers?
the closest thing to that is gemini
Gemini code assist is a thing - 19 a month
That's enterprise * code assist
oh
Nah the one I mentioned iz
mhm
The complete silence from Google regarding Veo2 is slowly crushing my hopes. What started as excitement is now fading into disappointment.
I have a Google One Premium subscription, I would get extra features like Gemini in Gmail and Docs. I use Google in the Netherlands. I don't see any of the features. What am I doing wrong?
How do you land a job at Google labs I'm a dev tho
Mmmmmmmmmm careers.google.com and pray lmao
like put 1206 on comedy central roasts
just for curiosity, we WILL eventually get access to veo2, isn't it?
not a Sora situation, when it releases one year later when it's old and no longer useful?
Didn't I remember that you already had test permissions?
lun doesn’t
Day 18 of waiting for the loop to end
!
excuse me?
I sure wish
aaaaah so THAT is why you were trying to dm me
you wanted to buy my account, Cheng. Stop doing that.
An issue I keep noticing with Imagen:
it will ignore your prompt if it decides it's not the style that a scene should have. It keeps turning scifi, horror or fantasy scenes into drawings or cgi, when I ask for photos and real-life.
no alien organic cave, however
Photorealistic woman in tattered white armor, quirky expression, in alien tech mining cave. Shot with 2007 Samsung camera, Cree Alto 35mm film, dramatic chiaroscuro lighting.
I did with my prompt! ^
it's more of a spaceship tho
You can edit the background prompt
But if the realistic style is what you are looking for.
the more real the better
I think the sci-fi keys triggered cg animated styles instead realistic
yeah I've just noticed one thing now,
this was, "real life cctv footage low quality, blurry blur amateur of a tall creature that seems made entirely of spikey bones. on its shoulder rests a plain woman in a simple exotic dress."
absolutely perfect
however....
real life cctv footage low quality, blurry blur amateur of a tall creature that seems made entirely of real, stained spikey bones. the lowe part of its body is almost serpentine and it emerges from a rolling pile of bones and sinew. on its shoulder rests a plain woman in a simple exotic dress.
......all of a sudden, turns it into a drawing?
Why can't you do that - if you're willing to do it, you can, and if you're not willing, that's your right. Or is there another reason?
Unfortunately I can't generate Whisk and VideoFX 
I just repeat what I want over and over in the prompt
and then it changes it. Like if it made a illustration when I want a photo, I ask for photorealistic photograph photo real image lifelike realistic photograph, basically every kind of word similar and sometimes twice. Usually works 
Yeah it's a good idea
https://labs.google.com/jules/waitlist/success waiting for jules
If you have access to google AI Labs you want to try ImageFX. The new version of Googles AI image generation model might be better than most models you have used in the past.
SOURCES:
- Google AI Labs 👉 https://labs.google/
- Image FX 👉 https://aitestkitchen.withgoogle.com/tools/image-fx
- Join The Out Community & Learn to rank our website #1 👉...
Whether you’re making AI stickers, refining your photos, or generating unique images from prompts, Whisk’s features are easy to use and endlessly customizable.
🎯 In this video, you’ll learn:
• How to create stunning stickers with Google Whisk
• Easy image edits using prompt-based refinements
• How to generate scenes, change styles, and combine...
Her is my blog post for using Whisk: https://wikifiction.blogspot.com/2025/01/whisk.html
hey, whats up, out of curiousity is this server composed of google or random people interested in google experiments?
does anybody know when veo 2 is coming out?
Ty
Hey all -- does anyone know of a centralized repository for prompting Gemini Gems? Besides the one that they give us, and thhe ones in AI Studio.
Hi everyone
Yay (cries. thanks but i really want veo 2 😭 )
"Cincinnati AI LLC now has access to Imagen 3 Editing and Imagen 3 Few Shot Customization on Vertex AI GA features!"
i feel like ill win the mega millions before i get access lmao
I feel this to my core.
?
?
Can Veo be accessed through Vertex?
Hey all :)
Local language channel not found
I want to create some GIFs for web3 projects, but I can't find the local language. Very inconvenient.
Question because I'm curious... I'm most definitely artistically challenged when it comes to traditional art. I've found there are surprising benefits to ImageFX and others, like expansion of vocabulary, learning new concepts, and discovering the creative process can be positive. Anyone else having similar experiences?
i'm sure plenty here do
Definitely with you!💯
Veo 2 is accesible through Vertex but a google rep has to allow it for your company, so if you sign up for google cloud you'll need to get them to activate it for you by contacting them. Are you building an app with Vertex API?
Over the holiday, I made a couple of music videos from a random MusicFX DJ jam using VideoFX (Veo2) outputs... here is the first part: https://www.youtube.com/watch?v=DeTOidSgZws
Starts a little slow, but skip to ~2:00 if you have a short attention span 🙂
I made this music videos using a couple of AI models and tools my team developed this year. The music was made in a single real-time jam with labs.google/musicfx and the video clips generated with Veo2 and then synced with the music.
The process of was very fun and hands-on -- many small creative decisions along the way, and powerful tools that...
Yes
any idea when Veo2 will be released? rough timeline?
fr, also doesn't work for me
😦
My friend said the problem ai can’t do close ups like a cat nose, I wanted to prove them wrong. Now I’ll never know
Nobody nose 
there's no way to do an image prompt yet with imagen 3?
could u give gemini a cat nose pic and then have it gen a similar image?
seems like it doesn't do the lens distortion enough
kinda got it to work "fish eye lense effect is very strong. a photo of the nose of a cute kitten ."
off topic, but anybody knows of a good lip sync tool for videos?
Can someone make "Gaming livestream of playing Dota 2, with facecam"?
thank you
Its too beautifull
like u
kinda got it to work "fish eye lense effect is very strong. a photo of the nose of a cute kitten ."
So .. I did this as it was presented to see what would happen . And I think at some point, I've managed to make Gemini snarky. First response wasn't totally unexpected. But the "have a right deadline".....?
Imagen 3 did it just fine via Gemini. Once the AI decided to cooperate.
ooo who let you in here?
Technically but I can't find out how- its not in the model garden (For me at least)
I've been in here for ages 💀💀
14th of December
And for your information... You did 😭
thats like too long ago for my memory lmao. garlic is the toast yall, shes super smart- no joke. plus she knows how to use excel, which is amazingggggggg.
😭
hey
Heya
they secretly released native image and audio generation with 2.0 in Vertex 👀
Saw that, But i don't think it's rolled out to everyone
yeah
Some tools are not available in my country and I want to use them.
Will they be available soon ?
how can I get access to veo-2 ? I have made over 10K videos with runway.ml and I wish to cross compare all of them with this one... Its been 2 weeks now since I applied for this and still didn't get back the email :((((
as of now we don’t know when they’ll accept requests again, but video requests are done in #flow-videos・📹 with @subtle vessel being one of them
you’ll know when it happens
correct
thanks... I am mainly interested in 1st part of your message reply, but good to know that there are people here that can give us a sneak peak through a personalized prompt (Hope i understood it correctly)
yes
and another question: why is there no more image edit (inpainting) when I use ImageFX??
ah! Imagen doesn't know the guira guira bird
Una imagen para un canal de Youtube enfocado a contar cuentos para niños
When google veo2 is going to be released please
Todo el mundo tiene talento. Las personas tienen el mismo valor pero no la misma habilidad. Algunas personas parecen ser bendecidas con una multitud de talentos. La mayoría tiene menos habilidades, pero debes saber esto: todos tenemos algo que podemos hacer bien. Cada persona es capaz de hacer algo mejor que otras diez mil personas.
"La vuelta al mundo", André Henri Dargelas (1828-1906). André Henri Dargelas fue un retratista pintor francés nacido en Burdeos, el 11 de octubre de 1828 y especializado en escenas infantiles y ambientes familiares, en una especie de realismo romántico que ha caracterizado la práctica totalidad de su obra. Su realismo académico, se mantuvo alejado de las vanguardias existentes en esa época. Obtuvo el éxito en vida y expuso en el Salón de París a partir de 1857. Murió en Ecouen en junio de 1906.
The filter's so crazy. It allows Shao Kahn from Mortal Kombat but blocks Raiden.
lol, Raiden's just.... offensive somehow
I like that I can make nude women but I can't make a fully clothed samurai with a mask, by the way
@iron swan bro gtfo, go to #imagefx・📷
my bad, I thought this was the discussion and imagefx was for the images - will move there
All good lmao
do you guys know any other way to call this thing?
it seems that I can't refer to this as a bl ack h 0le
Dark star, abyss in space, cosmic abyss, spiralling cosmic abyss, gravitational singularity
idk what that is but i would love to fall into it
As best I can tell, whisk is just another way to generate a prompt for ImageFX. Gemini analyzes the input images and merges the resulting text to form a prompt to ImageFX.
In other words, which is better depends on how good you are at generating an appropriate prompt on your own.
great answer
Welcome to the Monolith...
Made with Google Veo 2 / Runway Act One / Elevenlabs / Udio / Premiere
Written, directed, edited, and performed by me, a 100% dunce.
best veo 2 movie ever
Nice
Did u made it?
Good evening everyone
So VideoFX is indeed superior to anything else I’ve tried, and I’ve tried an awful lot. What I would love to see is a real time animation without the need for mocap equipment or complex setup, or even not real time, just mocap.
Yea, there are a few solutions out there, but very few with the ability to automate facial meshing for emotion to face animation or crazy thought, whole body emotion capture.
Nvidia has some technology out there too, since late 21’ but the facial meshing needed is difficult even for the most advanced animators in the industry.
I have ZERO experience so I dedicated months trying to leverage their platform and eventually had to move on.
Is Google working on anything like this? If yes, I’d love to be your “normie” test case and have several use cases that would be an incredible case study, well, one of them not so sure would be appropriate for Google to discuss, but it would be cool to see if it’s possible
So capturing movement through video?, if that what you meant there is a lot of services offering that
There are several mediocre services offering that, but nothing that could be used at studio level without an insane amount of cleanup.
I’ve seen it done real time by one company, leveraging Nvidia but I can’t duplicate it as a normie with close to zero production experience. Also that same company said it wasn’t ready to trust it in a true production environment
Let me clarify… with custom avatars.
Custom characters, most non human
Because sure, you can use actorcore or meta human and it’s… eh. That’s not what I’m looking to do
Right that changes things. Only Nvidia i seen pull this off with a somewhat good quality, Google deepmind currently is preoccupied with VEO 2, Imagen 3 and Gemini, so i think its going to be a bit before we see anything of substance
Hello, I am new to This server and also new to Coding (don't even know Basic HTML) soo could anyone give me a Road map or some guidance? It would be really helpful
ah basically you can share all of your veo2, imagen 3/imagefx, and whisk generations in this server
Mostly in #flow-videos・📹 #imagefx・📷 #whisk-gallery・🖼 you can post it there
Video requests are also done in #flow-videos・📹 sometimes too if you don’t have access to veo2
@subtle vessel is one of, just ping him to make the video you want
You don't need to be technical for any of these tools and that is kind of the point. What are you looking for guidance on and what are you trying to accomplish?
I see, But I wanna learn em and Figured this Server would be a good place to ask this type of question
Guidance mainly as in what to do as a Beginner starting from 2025, and for what i am trying to accomplish is, it's just a thing I'd like to learn just to improve myself
@raw bison @outer wagon @strong thunder Hello, can I kindly get accepted into the Veo 2? I applied to the waitlist whenever it first came out and still haven't heard a response
i mean a lot of us been waiting... also applied the day it was announced
even applied before veo 2 was out lol, with the regular VideoFX
I get that but they can grant people access if they want. I have seen many people with access
Not only you 😂
You special or something?
@real trellis I see multiple people with no following or reason to get early access. I have channels with over 400K+ subscribers on YouTube, so yeah I expect to have it if they have it
And it doesn't hurt to reach out to ask the Google Labs Team for access.
nice man
nah this is right though, i was JS u ain't the only one
@rain warren Here you go trying to check me
Go look up that name and see the checkmark on the profile
i mean it's in ur profile.... 😭
I own Faceless YouTube Channels, I am not posting to my main channels
Faceless Documentary Channels. So don't act like you are smart tough guy
? all i said was "nice" man
no hard feelings, i'll be ur first subscriber if you'd like!
Nah I am going to call you out because you had the audacity to screenshot my profile and put it up here when that isn't even my main YouTube account.
It doesn't even show that linked to my discord. I checked settings
Acting like you did something is crazy
acting like having 400k would put u above anyone else is crazier... especially when it turned out you couldn't back it up
Mind your business next time. I'll reach out to whoever I want
i didn't stop u or say u couldn't
I can back it up, but I don't have to prove anything to you
They are faceless documentary channels. Who are you to check me?
i was just letting u know u ain't the only one. don't get caught up in ur feelings
You got caught in your feelings when I dm'd the Google Labs Team
Why doesn't changing the Aspect Ratio setting in whisk take effect
How do I change aspect ratio of the images I want to generate?
Hey
really into all the projects google is working on
too much interested and excited to test each and every
That is something I noticed myself.
there's currently a bug. we're working on fixing this 🙂
current work around for this bug when it happens:
- change that drop down to a different aspect ratio
- change it back to the aspect ratio you want
- it works now
currently this bug seems like it is getting triggered when editing an image from the yellow sidebar
May I ask about the daily limits for videofx? Or is that a bug too? Just curious on some information regarding that.
daily limits aren't a bug 🙂
I can't comment on the specifics that we have set there, but all you can do right now is try again later
@dry bronze ok fair enough. Just no real specific information was given on anything regarding the situation
oh... when you hit your daily limit, you should see a toast mentioning something about reaching your daily limit. is that not happening?
I'll share this feedback with the relevant folks
I stopped using it yesterday at approximately 12:17 pm eastern standard time.
@dry bronze
It did in fact happen, but the daily never restored back
Not sure about any toast being there just that headline below
Wait is someone having the same issue as me?
Possibility
I checked this morning and I still haven't gotten my daily back
That was about 6 hours ago tho so I'll go check
Yeah that's why I had to ask to see if a bug was the issue instead
Wait mines back
Yay I can finally finish my thing of my character
Then I'll have all 4 of them done
I have the same issue. Today, I only generated four videos in the morning and got a message saying I hit my daily limit. I'm not sure what to do next.
are trusted tester applications being accepted yet
I am so sick of this message, how is it still there after a year?
hi, does someone know when whisk will be available for another country besides the United States?
Don't know
This might seem like a stupid question but I've always wondered like what's the difference between image fx and whisk
I mean too me they do a mostly similar thing?
join the waitlist
Anyone?
I don't use imagefx but they seem kinda the same? Seems like imagefx is just used for generating realistic images
While Whisk is made primarily for generating art
Ah ok
Whisk = layers (think of it as adobe photoshop), while imagefx is one shot/prompt generator.
So would it be better to use whisk for like OC creation or imagefx?
For visual inspiration, I’d recommend using whisk, as it excels at brainstorming and generating ideas by combining various visual elements. On the other hand, imagefx is ideal for creating detailed character concepts from textual descriptions, offering more precise customization.
If your OC has ready designs, I would start with whisk to brainstorm new ideas, if you know what you want (precisely), go with imagefx. Imo
Actually I used Whisk to make pictures of the characters for the book I'm writing, so I really recommend it for OC creation.
Cuz you can keep changing the pictures and get the result u want
While maintain subject matter
For anyone considering Sora while we wait for Veo 2, from experience is not worth $200, very broken outputs currently. Hoping I get Veo 2 access soon to make a full comparison
Hey, I wanna ask something...
Does Veo2 work better with complex prompts or simple prompts?
in my experience so far with Veo2 - use text to image with a complex prompt to get the base starting image and styles you want, then reconsititute some of that prompt into the motion section
Motion doesn't need to be complex
The $200 plan is for power users that need o1 pro extensively. It’s not for general users and certainly not just to use Sora.
You get it included in $20 plan
You’re right but I needed extensive testing as Im building a video platform while they release Sora as API, $20 credits are not enough to test it properly
Many video models offer better results for half of $200, so I just wanted to share for people who were considering it, Also when Veo 2 releases, going by what other people have shared it seems like it will be the number 1
Yeah that makes sense then
Hey Jaime can I ask you a private question in a PM?
are videofx trusted tester applications being accepted yet
nope, no new official announcements or anything
ok
do yo uhave it?
no
ok
ok
hi, is it possible that my daily limit reset got stuck? for the past 4 days I'm receiving the generation limit message without being able to generate any videos with Veo2
Can I use google veo 2 from Bangladesh. When I tryed to login to it, it showed me that It is not aviable in my country. Can some body tell me the way to use it from Bangladesh for free
Sora has only 10.000 fast credits for te 200$ sub, and you can do roughly 100-150 shots with that (depending on the resolution and length of clips you want) Its from 5, 10, 15 or 20s in 720p, and 5s, 10s in 1080p. After your credits are gone (mostly after 1 day!!) it has unlimited slower generations, which is ok for the speed.
In my opinion, Sora is best used for remixing clips at the moment, means you can "enhance and upscale" shots from other AI Video Gens and can get some pretty realistic results. You might loose character consistency, so this usecase is limited.
Besides that, it has a re-cut function to trim and extend clips you alredy generated, which is really powerful.
But as said, the prompt coherence and physics are not there yet.
Outputs vary from "wow" to "what´s that???".
Precisely!
Oh I never thought on trying to upscale videos made with other gens
Thats what I mainly use it for at the moment for 😅
Hi guys I'm so happy to be here
Welcome, whats your favorite labs tool
Wondering if anyone else has run into the library images being extremely pixelated. I've tried reloading the tab and opening it on different ones, but the previewed images keep showing up pixelated.
Actually I can’t access it in my location yet 🥲 I’m here for updates
Which one you’d like to access and is it for creating? Or for work?
https://idx.dev says that i can use flutter to build an ios app
but i only see the android option
anyone know how to build ios apps here?
(no ios option visible)
When’s the eta for Veo 2 access being public
I’d love to create and see what’s out there, because I create my own workflows 🙂 and you?
guys... is 1080p max resolution for veo-2? or 4k?
Neither at the moment. Currently it runs at 720p
I know. But Google already have plans. Let them cook
i cant wait to get inside veo-2
i cancelled my plan with the other ones just to keep money ready for veo-2
Good investment, because I still haven't seen anyone close enough to veo levels just yet.
do you have access to it?
Yeah always have since it released
amazing 😎
Dear Veo-2 guys !! please add me to the list ❤️ 🫡
They will soon enough for everyone
I will post an example for you out of good generosity if you provide a prompt @fluid quiver
thank you so much .. just a sec
Ok
is there a character limit per prompt ?
Illustrate an ultra-wide, panoramic scene showing a breathtaking montage of humanity’s most important historical moments, each seamlessly transitioning into the next. On the left, a view of ancient Egypt with the pyramids in their prime, surrounded by vibrant marketplaces and regal structures. Transition to the Roman Empire in full glory with towering statues, forums, and bustling citizens. Moving right, capture the Renaissance with iconic artists painting frescoes, scientific inventions displayed, and vibrant city life. Follow with an industrial revolution scene filled with iron giants and smokestacks under a golden sky. Conclude with a vision of modern skyscrapers, spacecraft launching, and holographic displays symbolizing the future of humanity. Each era is rendered in cinematic realism with intricate details and vibrant colors.
sorry if its too long
Ok it's fine
Thanks man
No problem
I normally don't post vids here so take it fast before they delete it
Thank man... 👌
Working on an app thats mainly focused on workflows for Film, turns out most of the tools on labs are perfect for it so here I am 🙂
illustrate a horse running through blazing fires
What a man gotta do to get early access to jules?
apart from apply to the waitlist lol
can anyone answer this
bruh i thought the ping was someone answering me
Meow
Anyone know if Google Imagen2 is still available for use..?
That’s great
Through Vertex
y'all my daily listen today was about daily listen 😭
podcastception
the male host is sceptical about it, nobody tell him
I am having a really hard time figuring out what i did wrong. I got the email saying I can sign in and be a trusted tester for a few different things but when i sign in, it says "join the waitlist"
you sure it was from a google labs team member?
I mean i realize i missed that cut off date, but it was before that that I tried to go in
hmm
i guess i can just try signing up again idk
ok i just assumed that it would be sent to a botmail account and ignored unless it says UNSUBSCRIBE lol
oh yeah
or you could directly talk with @outer wagon about it
well at this point its way past the date it was sent so i will just signup again i guess
will ANYONE ever answer my question??
Well, there was this guy who had all the best AI news and everything and always linked me to cool AIs I hadn't heard of, and now he's kind of.... disappeared? Deleted the account? Idk
missed connection: Kai if you read this, dm
loll TIL my profile is used on the marketing site 😄
and I found out from someone pinging me with a screenshot in one of my servers
I don't think they can do that?
Actually now that I think of it I did give them permission to use my stuff in their marketing material :)
so yeah they are free to
Tried an experiment where I asked Gemini to create some sonnets on a theme of AI and creation, refining them both by hand, and with alternating advice from Perplexity Pro and Gemini. Had a synthetic voice from Eleven Labs to read one aloud (with video clips by VideoFX) and finally generated an audio overview in NotebookLM to comment on the poems.
This is an experiment where I asked Gemini to create some sonnets on a theme of my choice, refining them by hand, and with alternating advice from Perplexity Pro and Gemini, used a synthetic voice from Eleven Labs to read one aloud (with video clips by VideoFX Veo 2) and finally generated a audio overview in NotebookLM to comment on the poems.
Interestingly, Notebook LM did follow the direction to cite each stanza from the three sonnets and comment on them, although I had to rerun the prompt once as it hallucinated a different stanza in one of the poems.
Hi there 🙂 I was wondering if we have a FB group that I can join for Musik Fx
Daily listen is easily one of the best features google has released i na while
I just wish it was longer then 4 minutes
is the remix button at ImageFX My Library disappeared justfor me?
I'm on the waiting list on VEO 2 since the lunch, any chance to get an access?
nah prob months away tbh
yeah
wat
Day 3 of daily listen 🔥
It disappeared for me too
Hi there 😊
I am a bit out of the loop..sorry
But what's going on with Video FX...or any Video Gen from Google..I see people posting things..are they the chosen one's?
BRUHHHHHHHH THIS PEOPLE AINT GIVING US ACCESS BRO I WAS ON THE WAITLIST FROM DAY 1
Videofx is great
where feature for EU https://www.google.com/photos/editing/
OH yeah I am friends with Kai as well....I noticed he went MIA too
I’m new to the server, does anyone know how I can get google ai responses?
you mean gemini?
Like the thing where u type smth into google and instead of
The normal google search it puts in the ai response
i think it should be fully rolled out, i have the ai responses
How do u activate it tho?
He always found cool tools!
frfr
You may have to turn on experimental labs tools on your google account to use it, also I think it only works if you have a google account and are logged in
Idk how to turn it on tho
And I’m logged into my google account
I’m in the UK aswell
Maybe there is limitations to what google tools are available in the UK and that may be why you dont have that option? I know google releases to the US first and as they get permission by governments they release to other countries
I thought it said “currently, only the US and the UK can use this feature” or sum like that
Unless it’s releasing to US first then the UK etc etc
u have to enable it, it's a experimental feature
Hi folks, any rumors about VEO2 release date?
If i somehow magically get into veo2 early, what are the limitations, is there watermark and is there limitations to how many gens i can go for in a day?
No there’s no watermark, and you cant do videos of celebrities
For some reason I just imagined a human getting kidnapped by an alien or something and the human is mad so they decide to annoy the aliens by making a bunch of references they'll never understand
Could someone with veo2 access dm me maybe, i would love to get one or two videos generated just to get the feel of the physics of the generations? ❤️ thank you so much
Google Veo 2 is fantastic when it comes to physics!
X: https://x.com/jerrod_lew
Instagram: https://www.instagram.com/jerrod.lew/
LinkedIn: https://www.linkedin.com/in/jerrod-lew/
Thank you but i was looking for specific complicated motion, other ai's really struggle with it and its said that its not perfect here either but from the clips ive seen it seems way better
theres many that applied to the waitlist months ago and not approved. this is not how it works unfortunately
When will they give global access?? It's been a while people have been beta testing....
Not commercially.
In fast motion hands, faces may have some morphing.
i applied to the waitlist such a long time ago... i have turned a year older ever since i applied
same
Which prompt structure would be appropriate to obtain the best result using text-to-video in veo2?
Standing out as usual with ur pink name i see 😄 Thank you
How do you make an iOS app in https://idx.dev ???
iOS development isn't supported at the moment, they also removed iOS simulator
flutter is inherently cross platform so i guess you can make flutter app for web first then later adapt it to ios
k
hi
whats doj
O
i think it the
Department of Justice
they want google to sell chrome because of the search monoply thingy
lost my phone
Hi Everyone! Happy to be here. We're the researchers at NYU ITP (where runway was founded) & creators behind the Elon Simulator: https://x.com/timwangyc/status/1871362178021720460
o
hopfully dont get banned fr
Can't wait for VideoFX, I already tried ImageFX and it was insanely good!
??
Is there an ETA on when Whisk would be relased in Europe/Italy?
Using flutter you could yes :)
Flutter is... annoying
Did you guys see that Anychat version of Gemini that can VidChat with you while discussing uploaded images/content
Google is pulling off some insanely impressive AI feats at breakneck pace
That lil AI content detector theyve got on labs is pretty cool too and assuming they fail to have any competition in that area of AI expertise, is gonna be vastly important to the future of probably all of society within 2 years max I would bet
are you still waiting for veo 2 access
13
14
1
yes
✅
unfortunately is too slow (free ones)
thanks! will look into it, if u have any link feel free to dm if so ❤️
do you think google have to sell chrome? ( the DOJ thing )
11
12
1
no
❌
W
Hy
I have documented some of interesting findings from the Lab in a Medium Post:
https://ankitaguha256.medium.com/multimodal-models-with-google-deepmind-6bff963df787
I have waited for veo2 for almost 3 weeks...
they only really accepted requests as soon as veo 2 was announced, but after that they stopped and went radio silent
there's other tools out there with more features....
Better tools than what? I use minimax. I am impressed with the graphics, but feel like I am still missing out on something BETTER?
Should VideoFX be able to create an AI scene from a VIDEO not just a photo?
Everyone uses minmax.... But there are others
Plz recommend some. But in my opinion Kling, Luma, and thoses all have MORE TOOLS but are far behind Minimax graphics. So m very interested to know what's better than Mmax
Luma just released Ray 2 yesterday, which is almost as good as Veo 2 except only 720p maximum resolution, although their servers got overloaded and outputs are slow right now even for the unlimited plan users.
proton is also a good one, most free vpns are not safe, proton is saf(er) then the other "free"ones
yeah, i was gonna say that lol. proton is also unlimited and has free countrys but only like has 3, but has alot of free servers. proton is safer then the other ones and is a trusted company. mullvad also is a great one but is paid
windscribe is also a saf(er) one but does have a limit
but does have alot of servers
yeah, if your not using for privacy use whatever, but you have to remember that everything is logged and not private. if your using a vpn for restricted places, use mullvad or proton since it helps with that.
proton is a saf(er) one and they have not sent any logs (they don't have any) to the police (YET). they did send some stuff to the police from the mailing service (it was not encrypted)
mullvad was also raided by the police and they asked for the logs but they had none
also mullvad has a free proxy (only for firefox tho)
a proxy is like a vpn but only on browser and less private but still safer (depending in company)
they have a ton of countrys on the mullvad proxy
yeppers
If you are using a free VPN, a public VPN, or even an IDC server IP, and you are located in EU countries, China, or Hong Kong, you might face implicit restrictions from OpenAI and other websites.
It's not about "data security" being important. What's most important is that the service provider (like Google) knows you are using a VPN, and they will take covert actions against you.
Hi @strong thunder , sorry for the ping. I sent you a PM two weeks ago. I can't provide more details here, but when you have time, please take a look. Thank you!
oh btw i found out u just have to use the vpn to enter, then u can set it off and keep using it GG
minimax is far behind sora "graphics". not saying sora is better at motion etc, just talking image quality at 1080p
How can I turn off Gemini for Workspace. I have opted out for my entire organization and it is still showing in my inbox and when I compose an email.
Sora supposedly was trained on a lot of Unreal Engine data, which is why everything looks super sharp, and often videos look 3D rendered and water always looks like mercury metal liquid
Interesting about the UE stuff, although I have to disagree my videos look very photorealistic and not 3d rendered for the majority of them unless prompted otherwise
3D rendered doesn't necessarily mean "they look like plastic", for example if you look at fan made re mastered UE games on youtube, they look hyper real, or very close to life. This means that training on UE made the outputs of Sora look very photorealistic, very similar to the idea of training a LoRA for Flux to look less plastic by training on very low quality jpgs, but training in low steps just to eliminating the 3d look
I see, i actually prefer a more clean look maybe a lil bit like 5% 3d feel compared to the grainy stuff to make it look like amateur pics/vids. just depends on what project you want to work on. I agree it looks like they trained it on that kind of stuff a lot
VEO2 though is much better in motion, coherence etc
only quality of sora above the rest is the image quality. although its been said VEO2 can do 4k so lets see
I agree with you, Sora looks super sharp, though since I have worked as a DP, Sora often looks overly processed, almost HDR. However if Sora would actually follow prompts 😂 it would be amazing combined with that quality
Also I hope they fix the lighting, currently Sora often looks like its centered lit directly by the sun, the lighting is very harsh. Veo 2 in comparison has an amazing understanding of lights and shadows
Completely agree 👍
I did not know that whisk and imagefx had a Daily limit, but, i reached It ahah
after how many generations?
Dunno ahah
The text was even bugging out with <Headers> link <etc>
It's already mid-January... Google didn't keep their word.
what was their word?
At the beginning of January, veo2 was announced.
https://youtu.be/RRBk9nGcNuo?si=nH-egjIwbznrR7d1
Made with Ai except lyrics ❤️🙂
"Echoes of You" is an emotional song straight from the heart of a lonely, overthinking teenager. It’s a raw expression of unrequited love, regret, and longing for connection. If you've ever felt lost in your feelings and unable to let go, this song is for you.
Hashtags:
#EchoesOfYou #EmotionalSong #HeartfeltMusic #UnrequitedLove #OriginalSong #...
why is there slowmode here now
-# This user is wanted by a United States Federal Agency, the KGB, +9 more.
-# If you have any information on the user's wearabouts, please click here.
is anyone else getting 'internal error' when using ai studio?
every now and then. once i hit a certain threshold with the conversation it gets more pronounced
its weird because it was my first prompt of the day haha
any idea when 2.0 will allow pdfs, though?
only reason I still use studio rather than my actual gemini is because of that
it's usually very slow in the morning... as a lot of people tend to use ai at this time lol
pretty sure it does
mine only allows images
it should allow pdfs as far as i'm concerned. pretty sure i've uploaded pdfs before
oh you mean directly on gemini. not sure. ai studio will though
so yeah I've been using the ai studio one for analysing pdfs
yeah I just prefer gemini haha
because I dont run into those annoying errors and limits
go under settings and enable google workplace then try uploading the pdf to ur google drive
thnx will try that
I dont see google workplace
its under extensions
got it!
yeah not too sure it will work but it's worth a shot
yeah I switched it on but still only shows the image icon in 2.0
its funny that this is the same issue I have with o1
they are great models but both dont let you upload pdfs
sometimes I just copy and paste the text from the pdfs, depending how long it is that works
thats as far as i got so far lol
i just use my own apps tbh.... made a discord bot that can do all that already
based on gemini's API
what the heck lol
you wanna see something REALLY funny tho?
BRUH lol
tbf it did make me laugh out loud, I had to screenshot it
restart the app and make a new chat
ok I will try your exact prompt, one sec
make sure workplace is enabled for sure
weird that it doesnt with advanced
u would think "advanced" would mean it actually has more capabilities
yeah it does make me wonder if it's too expensive or something
I mean for them to allow
Paid $100 for Runways new Frames and I am here to announce that Google Labs Imagen 3 is way better, and its free 😍
on my third week of waiting
It's important to remember that this is a community discord and not a deciding making arena. It would be like going to a World of Warcraft forum to say you didn't RNG some loot even though you really wanted it for a long time and other players have it. .The roles of community managers and bug fixers don't allow them to do anything, even if they all wept at countless stirring essays each of us wrote about our collective wonder and joy.
Workplaces can take a while to turn gears, get approval, wait for another team to finish their pieces before the next phase.. it's literally not about us.
good companies listen to feedback and messages get transmitted upwards hopefully though 👍 Ive not been complaining myself but I can't wait to have access
Yeah, I'm also looking forward to access. There's no guarantee of inclusion though. They likely allowed in as many people as necessary for their purposes, which in Labs often means research, testing refining,, gathering feedback.
And the labs we can access are still interesting. Plus the wide world of readily available tools beyond Google.
I don't know. I'm bummed scrolling back that channels have people filling space with complaint-demands. Then they bounce, never contributing anything else. That's not why I use discord anyway.
Not all 5 digit membership discords can have so singular a vision as a Morbius movie, I suppose.
I mean public access, Im not even waiting on getting it early. I hope they are near the public access now. It certainly looks more ready than most out there lol. And given Its google I'm sure they have no issues getting enough servers. I'm ready to pay whatever subscription.
Yeah, given that Gemini explicitly avoids generating people and such. encouraging GeminiPro subscription to do it. From what I've seen it's pretty close to ImageFX, plus some LLM tricks.
I'm sure the monetization to better sustain itself is a goal. In a way a subscription service would be nice because it formalizes an agreement. As is, I feel like a stranger looking to consume resources in exchange for my prompts and the results content.
squints suspiciously
just use other ai that make videos, there's plenty and can do more than veo2
Whoever has Veo 2 access at least I hope they keep sharing the amazing results with us 😀
tiktok ban tomorow!
yall got veo 2 access yet??
18
19
1
no
✅
Free?
Hi guys
Hello my friend
Came early
honestly not even complaining tiktok is gone
been brain rotting people for way too long now
L
I hate how obnoxious the filter is
When is Gemini 2 Pro suppose to be released? Still waiting ot access Veo 2 🙂
all these fun toys, yet no acces 🔥
weird question
how do i make an image look like it’s shot on iphone, without mentioning that it’s shot on iphone?
“Make an image which was taken on 48mp”
mp means megapixels btw
How long is waitlist for veo 2
well, tried, but it didn’t felt like it’s shot on an iphone
this is an example btw
made this on accident with imagen 3
How about instead tell it that it was shot on a 24mm wide lens
Hmm now add that but with 48mp in the prompt too
eh…it’s not working out
How about adding “old snapchat look” into the prompt
i think this is good, ig
You could add medium detail
eh…
honestly the last one feels like it’s shot on iphone
here's the prompt btw: photo of a cat, taken with a 8MP camera, 24mm wide lens , old snapchat look, medium detail
i’m going to bed so have fun
Yes
I need Veo 2 access
Join the waitlist
did that LOOOOONNNG TIME
Yeah we haven’t gotten any confirmation as to what’s going to happen rn
I am patiently waiting. But anyone have any ideas on the Monthly Unlimited price yet?
I put in just your sentence . First round it loved women on benches.
Round two brought college promotional photos and a fisheye.
SD, MJ, and VQGAN for comparison
And their round two.
whats up guys
what do you mean?
wh1t3list bro
Does anyone know how long it takes to approve the video?
for create video
Just a few more minutes.
I already did it a week ago and I still can't use it, do you know what it could be?
as I understand it, they are not accepting any new people to it, we are all waiting for it to release to the public to be able to use it
):
Does anyone use Chat GPT for prompting images in Image FX? Im trying to get more realistic images out of IMage FX
Tried it with Midjourney and got mixed results.
dawg i signed up when VEO 1 was announced and still no access
each model has different prompting style and also you need to give instructions to the gpt corresponding to what the models responds to. you cant just come to gpt and say "give me midjourney prompts" or image fx, give it some guidelines, some example prompts that worked good etc
My dream is about unity amongst humanity in all realms of sectors, finance, community, understanding of all ethnicities and kinds and peace to all. Take a listen. Could this potentially be something that could be a reality in the future?
