#general・🧺
1 messages · Page 32 of 1
Thanks @outer wagon
anyone know the difference between Pro and Ultra google plans for token ammounts?
Not right off good ?
Im new here, tried to use project genie and got this. anyone else? Is it down right now or am i doing something wrong perhaps?
i tried refreeshing and clicking all tabs
me too
gogle
Hello, as stated in #start-here・📍 , “Currently, our team can support engagement in English; the discussion in this server will largely remain in English for now. Thank you!
1000 vs 25000
Where do you live?
canada i decied i cant go with it because it cancels my pixel 10 plan no discount
going to subscibe to claudecode for the time being until my pixel 10 plan expires and then use both
This is the way. No single set of tools does everything best.
working on my own design Multi Angent node structure with a auto populate workflow of agents to build stuff currently saves a lot of tokens. I found creating secraterys and full office structure helps with local LLMs however the larger context windows I need good LLM agents to make decisions to push back. Gives me a full range of design with vibe coding and designs.
can i get free trial of genie 3
genie 3 is behind a paywall because its not flushed out yet, latency, limited controls and inaccuret prompts. I am not sure but I speculate it takes a lot of resources to run that 30 second demo as its only 30 seconds. Wont be released to other tiers until they probaby find a way to get the costs down.
BYODC
Bring Your Own Data Center
i been trying to add API keys to Antigravity? is this not possiable?
Like a Gemini API key even?
i was interested until i read the languages
we're on completely different realms
my native tongue is C# but i also know Java, C++, C, Rust, Js, Lua and Python
In my native land of C#, we call those 'NullReferenceExceptions.' It’s the same crime, just with a much longer name and a better-dressed judge.
Final question: If you know all these languages, why did it take you three hours to center a <div> in JS?
british columbia canada. is it just US rn? if so anyone aware when its supposed to be available in canada?
also, ive been using higgsfield, i got 8k assets created infact over past 2 months. used base44, but anyone have complimentary cutting edge platforms/ai stuff besides those that is there fav/unique and powerful application on the rise??
Google doesn't decide. They have to go through an approval process. This keeps coming up, so I looked up how long this took for different parts of the world, and a pattern emerged, so I made an infographic in Nano Banana:
Get a VPN and address and subscribe down it the USA. 🥷
oo
LOL thats a very poor decision and probaby aginst many TOS
what about like the most poppin AI creation communities for yall who are super into it? Like i said ive made 8kish assets and i dont know many other human beings as into it as I am and im interested in meeting like minded peeps pushing the limits
Bunch of people in here have tried that, and it rarely works. If they know where you live, using a VPN doesn't change that.

I know a way but rather not share bad decisions and would probably cost double what its worth to do. However easy when you have family, a FRP tunnel or Oracle server and easy address to use.
Anyways Arena.ai got a large update. Looks like google in in 2nd place fore image gen again. OpenAI took the lead just recentily. however these are models not released to even the USA
Oracle Cloud Compute (ARM): Up to 4 ARM Ampere A1 Compute instances with a total of 24 GB of RAM and 4 OCPUs. This is significantly more powerful than any other free tier. Google Cloud Platform (GCP) provides a steady, low-resource "Always Free" tier suitable for small scripts, bots, or light web servers. You can take advantage of to help with compute and stable IP address.
Hello, i want access please im using ia everyday thanks you
can you solve this bug?
Hi there, I'm going to forward your message to #whisk-general・💭 which is for questions and troubleshooting!
reminder, general is not the place to ask questions about products as we have #flow-general・🌊 , #whisk-general・💭 , #jules-general・🦑 , #opal-general・✨ and many more channels where your product specific questions can be asked. Thank you
You've reached your Nano Banana Pro limit. Images will be created with the previous version until your limit resets on 3 Feb, 16:26.
Why is this showing when Im a subscriber Pro
created under 10 images when Im allocated 100
or 1000 for NB fast
Did something happen with AI studio? I vibecoded a tool a while ago but it became unuseable because I get hit by API limits way faster than before
as of this week I been hitting my API limits on antigravity in 1 hour instead of never with my pro plan
something is taking more tokens
The models ig
I use Google Flash 3.0 the most it seems to max out super fast though
Hello @fossil yacht & @next scarab as stated in #start-here・📍, “Currently, our team can support engagement in English; the discussion in this server will largely remain in English for now. Thank you.
yo chat i need some tips to become a pro at coding. What should i do?
hey i am Adii from india i am searching for aiml certification course, any suggestions?
Uhh
🥒
wow
/pickle "flappy_bird_2"
Hello! I'm getting an issue with a download for my testing program. It's giving me a 403 error.
I've yet to be added to the disco channel so I can't ask there
Seriously Google, I can't upload a document to Google Drive so that my coworker has access to it? Automation becomes useless because of this (Flows).
vibecoded .. new word? I like.
WTFrick is happening with google flow??
I cannot using my full credit to generate any video, and in this february my credits is got reset. my old credit is gone!!
Are you guys in SCAMMING business right now?
this is so dissappointed, at least let my credits there if the machine is still getting fix, why are you guys still STEALING customer CREDITS!?
im using veo3 with an API key and trying to create an ai transition. can someone help me? it sucks right now, and i dont know if its my prompt or wtv
i know this is for google labs but i dont know where else to ask and i do not want to switch off of veo
Higgsfield is one of the fastest-growing GenAI platforms, built with one clear mission: to be the most creator-friendly AI video ecosystem in the world. With Kling 3.0 officially coming soon, the AI video community is already buzzing — and in this video, we break down everything we know so far, plus what creators are realistically expecting ne...
What is this?!?! ._.
try in #flow-general・🌊 🙂
Does anyone know the global release date for genie 3
Can we have a private chat?
Sure
I can't send you...
Could you please send me request?
Hello as stated in #start-here・📍 , “Currently, our team can support engagement in English; the discussion in this server will largely remain in English for now. Thank you.
YouTube is hitting AI creators hard rn. Every bit of help is welcome. If anyone has any sort of direct contact with Youtube tean, your help will be highly appreciated.
If you run a YouTube channel, you could be next tomorrow — this is happening indiscriminately. Today it’s me, tomorrow it could be you. Let’s stand up for each other. AI creators who make original, non-slop content need to support one another, especially in times like this.
My channel name was "SithDaddy"
Also, its kind of crazy that with one hand Google is developing Veo3, NBP and all other cool tools, and with its other hand its terminating AI creators' YT channels left and right without any trial or investigation.
If you're creating Star Wars content, it's likely a copyright issue
No, if it's a copyright issue, they will just make certan videos unavalible, they dont kill the channel for that. Besides, there are hundrets of SW channels and they are fine, nobody closing them. And I was not making purely SW videos as well.
YT has a 3-strike system for copyright infringement. And even though all my videos were parodies, so 'fair use' implied, I never got any strikes or anything.
I started making political videos recently and my guess is that this is what triggered them.
Cos I made a music video about Donald Trump like 2 days before my channel was axed.
That's likely as well
Wsp is antigravity bugging out for anyone else? Keeps hitting me with the agent execution error
hi
Hi, is there a place to specially discuss Antigravity?
Think it's usa only
will it be available in other countries after some time
Highly likely, but the wait is bound to be several months if not longer
Hey there, I don't see the disco channel
we recommend the Devs server for now, they have a vibe-coding channel which has been the unofficial Antigravity space: https://discord.gg/JahnXh4N
epic. thanks!
hello friend , I am new here .
😑
How ok is google with claw using the antigravity limits
Labs does not have an official date for expansion of Genie 3. we'll ping the Project Genie role when we have updates to share!
hi, i just heard about this website
i wanted to try it, but i keep getting this error
what is this?
hey Salam! Project Genie was rolled out last week to Google AI Ultra subscribers who are located in the U.S. only, and are 18 years or older. please note that there are certain restrictions, as Genie is not available to Google AI Ultra for Business users (see article below for more information). If you meet the above requirements, head to https://labs.google/projectgenie to get started.
if you meet all of the above criteria but are experiencing issues when you access Genie, please reach out to Google One support in case there's an issue with your account (we can't look directly at accounts here on Discord!)
📌 Resources
oh, so when it will be able to work in europe?
and i will check it out
thanks 😁
we don't have any official updates at this time for the expansion of Project Genie, but we'll ping the role when we have some to share! you can grab the Project Genie role in id:customize
I just wanted to try Genie, but received a 404 https://labs.google/fx/projectgenie/404 .... I guess, because I am from Europe, right? It would be nice if the error message would be more clear 🙂
that's correct - Project Genie is available in the U.S. for Ultra subscribers that are 18+. thanks for this feedback 🙏 I'll send it back to the team!
thanks a lot
i will be stay in touch
and what i heard that i can create video games and music stuff very fast and easy in this website, is that correct?
because i am working on mini game right now, and it took me 5 months and i still not finish it yet
check out some community examples that were shared in #project-genie・🌎! you'll be able to get a feel for the outputs there
thanks, i will look at that
@outer wagon disco link posted yet here?
ayy thanks man
this is not an official release document or date from Labs, thank you!
Hii, is "Gemini in Chrome"'s Auto-browse feature part of Gemini? I have some bugs and things I'd like to share but I'm not sure if I should share it here or in the Gemini server, given that Auto-browse seems to be powered by Project Mariner (which is a part of Google Labs)
oh man, i got excited for a second 😅
thanks tho
Hi
Hello
Dang bruh we have to wait untill pubic release
Talk to your politicians. If it was up to Google, they'd release it to everyone.
I doubt they listen to normal people
Yk I just get the ultra plan to see if this genie is good
They certainly don't here in the US, but we're lucky that they haven't impeded AI as much as other countries have, at least not yet.
This is a pretty huge issue, so if enough people make noise about it, and tell politicians they won't vote for them, then they'll either listen, or lose their seat next election.
Hello, what do you expect to come with the next Veo model?
Damn fr I’m in Wisconsin but they still do it anyway
Hasn't been anyone on a ballot I could vote for in states I've lived in years. Just ballot measures and whatnot. We'll see how long our luck holds on AI issues. Broken clocks and all that.
Any ideas like I have a lot of ideas but not enough resources to do it I just want to save the planet by implementing worker drones and other things to help old people who family left or died
It'll gut employment, but it'll definitely start lowering healthcare costs within a decade, and domestic robots will be useful and affordable.
You’re a nice person I’m very glad to meet you
Hello.
Hi, is anyone there having issue in the Google labs?
No why?
Because I can't enter there, it's an error 404
Hello!
wasap people
Site loads fine for me.
Have you tried clearing your cache or using a different browser?
Hi!
is this logo quirk intentional cuz its lwk tuff
hi everyone
Hi everyone
HI
Hi
halloooo
I got problem with Genie 3
It gave me Error 403 screen
As an Indonesian, only Americans enter genie 3
So Google Labs hates Indonesians.
Hi everyone! What are you building today?
No, Google hates the half of the world on that point
I don't mean to say the America is the half of the world
VEO 3 is finally available for Indonesia on 2025, but Project Genie? Unknown
So, Project Genie is only available in a few countries?
I hope Google fixed the Error 404 on the Genie 3 and available for Indonesia and others
But no subscription at all, free 200 credits renew monthly
I guess
But you need to waste your money to get credits idk
Google is definitely Mr. Krabs
Oh, you need Google AI Ultra AND a US google account. I do have Ultra, but my acc is brazillian... i pay money and can't benefit, great...
where is #disco-announcements channel ?
First thing i see after joining the server btw
what?
i just joined this server and that message is the first thing i see lmao.
India.
in India where/
Mumbai, Maharashtra.
i'm studying for now. how about you?
I AM A HIGH SCHOOL STUDENT
that's great!
IN 11TH STANDARD
oh a young explorer! that's great!
hey hey
hop on the labs-lounge if you want to voice chat!
got a couple folks in the chat here
Hello everyone i software developer and here to explore AI tools
hello
"Hi team - I'm generating videos in Veo 3.1 with prompts that explicitly state 'NO SOUND EFFECTS OR MUSIC OR NOISE' and 'SILENT VIDEO - NO AUDIO GENERATION' at the end, but every single video is still generating with synchronized audio. I've tried multiple prompt variations including 'visual only, completely silent, no audio track' but the audio keeps being generated anyway. Is there a specific prompt keyword, parameter, or setting in Flow that will force Veo 3.1 to generate genuinely silent/mute videos without any audio track at all? Or is audio generation currently mandatory in Veo 3.1 even when explicitly requesting silence? Any workarounds appreciated!"
Can you export the worlds you make with genie? I would love to try people's ideas lol
We need a new tool called Slop - it will auto generate a prompt and random ai slop on a single press of button - 1 click 1 slop. Comes with auto upload&publish to youtube shorts.
Only one.
It's not a world.
It's automatic with the 3.1 models.
*in Flow.
It is brutal. It is easy enough to edit it out but that is another layer of work that isn’t needed.
How hard can it be to have a toggle prior to exporting to turn the music/sound off? 😕
Shouldn't be. The models in the API have versions with and without audio, with the non-audio version being a good deal cheaper, so it's not a functionality issue, but a choice in how they build the Flow web app.
100% agree. I even used the same prompt in the Veo 2 with no sound and it says " Not compatible. Please use Veo3.1 - Fast" 💩
That's something else, I think. Ingredients don't work with Veo 2.
I was just using text to video
Weird. That's definitely not how it's supposed to work.
interesting
I can not wait for Material 3 Exprssive Web full drop !! really exiceted to work with it
i want to be staff google
After generating screens via the Stitch MCP API (generate_screen_from_text), the screen cards in the project dashboard display
perpetual gray loading placeholders instead of rendered thumbnails.
?
hi, so im facing an issue where when i download all my images on whisk it seems to crash idk what to do
#whisk-general・💭 would be better for you. Check out #1319422603001200731 too
When will Genie 3 release in Europe?And How much wil it cost?
hi
When i manage to make this.
Hiii
hello
Hi there, head on over to #flow-tips-n-tricks・✏️ for prompting help. 
@noble quartz sup
ayee

what are you doing here?
Hi all who is of what sort of occurrences or tasks? I can imagine that people are of some sort of assignmeng however beyond that maybe temporary and maybe simply of some sort of appropriated relevencies and it seems as though usually people are attending their own matters not crossfunctions
hello
Same question
sounds like a solid idea 👌
if you want to pull ethereum data directly from the chain, the first thing to know is that everything comes from blocks, transactions, and event logs - no magic beyond that.
use Alchemy or Infura to get node access
If I process data directly from an Ethereum node, how quickly can I see that data after it’s impact on the exchange ?
Also from a trading perspective — if I detect something like a large USDT transfer (for example $1B moving on Ethereum), how long does it typically take before that activity starts affecting exchanges or market price?
Is the market reaction usually immediate, or does the impact happen gradually over time?
if you’re reading directly from an ethereum node, you’ll see activity basically as soon as it hits the mempool (pre-confirmation) or within 12 seconds once it’s mined into a block. so from a raw data standpoint, you’re very close to real time.
from a trading / market impact perspective though, it’s a bit different:
Hello @frigid quest, 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.
reminder, general is not the place to ask questions about products as we have #flow-general・🌊 , #whisk-general・💭 , #jules-general・🦑 , #opal-general・✨ and many more channels where your product specific questions can be asked. Thank you
love the palestinian flag bro

Hello everybody my name is Cristián I am from Buenos Aires.
Yes'm?
How do people make these Ai games then
They're virtual spaces you can just walk around, using Genie 3, if you're in the U.S. and have an Ultra account.
I’m sure you can do it using a pc
Understand but I'm confused why they made this #general・🧺 then? and we also have specific products general... confused 😕
@shut garnet you are burning tokens, generate small responses
It's not a video game, where the area was built in advance and just has to be displayed. It's generated on the fly, like Veo does, but a LOT faster and more complex.
Do what now?
I think it's just 3D VR
general is for general talk such as asking where to go for a product or some general chatter around labs products, we recommend going to the individual channels as each team looks at those more than general chat. you will get people who use that specific product and know the product better in those channels than you will in general. we like to keep the server clean and tidy as well
When it’s coming to uk
Months, usually. Depends on your laws.
https://labs.google/projectgenie - this might help?
The link just takes me back here
most #labs-announcements・📣 products are fancy rather than productive ⚠️
Respect Google devs
Anyone knows how I can generate millions of volts?
Rub your feet on old carpet for thousands of years?
Well, hello there. How is it going?
👀
Lick ~/>2,223,000 9-volt batteries at the same time?
Frontier Modal??!
oh genie 3 is gonna be part of youtube thingy
Didn't even notice that, hahahaah
I mean... they should, haha
Imagine what we could make this space be... oy
i found this
Hey friends. I was wondering. Could you build an OpenClaw experience within Gemini?
It’s to many people on it
I want tô create an Agent with the same ease as I create a Gem.
What precisely do you mean by "OpenClaw experience"?
thats not as easy as you think you need to have all keys of the devices and softwares and its realy complecated
I want the model tô have a customizable personality and name. And for It to acquire skills and Interact with other models.
I dont Trust tô install OpenClaw
But If I could do It on Gemini, It would be nice.
Woo tried the clawbot?
yay!!
Someday, but I think he's referring to long term plans here, when all these models and tools are more mature.
I hope YouTube doesn't turn into something similar to the Sora app.
I don't think that's likely. There are SO many channels there making content people actually like. People aren't going to stop watching that and switch over to spammy stuff just because it's made with AI (otherwise they'd already be on TikTok, haha).
ello
is it possible to get banned from google
Hello 👋
Websites can get "delisted" from search, if that's what you mean.
i just found out what was slowing my computer
steam
now i can send messages without it taking a decade
Mine runs on internal combustion...
im talking abt the app steam
Haha, I know. Come on now, haha
i had to stop it on task manager
hey there nice to meet ya
Hi
hey where do we go to download anything
has anyone been added to the Disco role?
email only says to join here and wait in the lobby (but there isn't one?)
Hi
hey
how are you
i'm alright
where are you from?
the US
Yes
Why?
cool
kul
Same
I also just filled out the form, waited here, and there was no response
awh
i guess we’ll be waiting
Has anyone joined the disco channel?
OK, wait, I hope you can feel a completely different product
I use USA VPN and it loaded
Hack google
i bet
Or Ask one of their employee to give you the whole model (illegallllyyy)
Even it's going to end up with incapability of running the model
Why i can't use genie 3?
Is the disco team there? Can it be activated as soon as possible?
Can't wait to use the product
in general how long after completing the form for Disco does it take to get access?
At present, the time is completely unknown. I thought I would fill out the form immediately
my antigravity keeps dying
Hey everyone,
I’m looking to build a more efficient system for my Facebook page. My goal is to catch viral/trending events as they happen and quickly turn them into high-engagement posts with "scroll-stopping" captions and matching AI-generated images (Or using the google image api to get the newest image matching with the caption)
Does anyone have a solid process, prompt, or automation stack (Zapier, Make, etc.) for this? Specifically, I’m looking for:
The Hook: How are you prompting AI to write captions that actually work on Facebook? (Short, punchy, curiosity-driven).
The "Viral" Filter: Is there a way you’re automatically feeding trending news/events into a LLM to generate these?
Image Synthesis: Any specific prompts for Midjourney or DALL-E to create images that depict the caption rather than just looking like generic stock photos?
The Stack: If you've automated this from RSS feed -> GPT -> Image Gen -> FB, what tools are you using to keep it seamless?
I’d love to hear your "secret sauce" for the prompts or any workflows that have helped you stay ahead of the curve without spending 5 hours a day on Canva. Thanks in advance!
Unless you have a little data center in your basement, you couldn't run Veo, much less Genie.
Most who apply won't. It's a beta. Tons of people apply. I applied for a bunch and only got into a few.
Wow, thank you for this information...I had no idea
So no point of paying for Ultra ya?
Is the friend disco channel open?
If you want to use Genie, you pay for Ultra because you're running it on their servers.
I don’t think they planning to put genie model for api use idk
I really want to experience when disco will be available for download
Yes, it won't wait until they go to work
Still 😭
Haha Disco
i can't use Genie 3 any solution ?
@chilly shadow @lone zealot @tepid skiff
@heady pumice
What about disco??
Hi
When will the download link for the Disco browser be released?
I also want to ask this, download the link to publish
Did you fill out today's form?
I filled it out about an hour after receiving the email, and now it's been 6 hours
yes
yes
I guess the Google staff wants to wait 24 hours to unify the statistics of people who fill in and open permissions
I got access but now the app doesnt start (crashes) and i dont have the role to access the channel to discuss that, any mod who could help?
How did you get permission? Is there a download link? Can you share? Thank you my friend
Received an email, check your spams maybe
Took 24h to get it, and no cannot share anything sorry
I haven't received the email yet, only real spam is in the spam folder
It is already in the E-mail.
How do you make videos longer than 8 second in Flow ?
Damnn bro, you got the early access 😭
When will Genie be free to the public?
why need genie ?
because after all, it is the next generation of ai.
i'm curious to see the capabilities genie has, that's all
it can remember environments, AND last for up to several minutes while maintaining terrain, and it's still in the early stages.
unfortunately, i'm completely broke to afford genie 3, so we just gotta wait.
i mean most of us are broke
i think...
someone ghost pinged me 👽
You can try Microsoft copilot gaming ai but that's only fit one game, and not like genie that has custom prompt
reveal yourself human 👽
Oh?
let me see
About your bio, what actually is consciousness
@crude moonI have not received the download link yet.
Wait for you to get approved by Google Labs team.
I have received the approval email, but I have not received the download link.
What did the E-Mail say?
Dont directly copy paste it anywhere, tell me if it says "Download Disco Installer"
This word
Thank you for your interest in Disco, a new Google Labs web browsing experiment.
You are part of the exclusive first wave selected for testing the prototype.
You are not IN yet, Wait some time for their Approval.
if you are in the e mail will state it informatively
I want to try DeepMind.
Does anyone have an account?
create me logo avatar
hey
Hi there, please DM @fringe nova and open a ticket so we can discuss this further and get it sorted for you. Thanks!
Hello
.
@frigid quest @lucid sigil Hello, as stated in #start-here・📍 , “Currently, our team can support engagement in English; the discussion in this server will largely remain in English for now. Thank you!
Chicken
Chicken banana
what is happening with Nano banana today?
Yoo
is discord dying tn
reminder, general is not the place to ask questions about products as we have #flow-general・🌊 , #whisk-general・💭 , #jules-general・🦑 , #opal-general・✨ and many more channels where your product specific questions can be asked. Thank you
how to fix this
Project Genie is available to Ultra subscribers located in the U.S. who are 18+ years or older. this error message could be for a variety of reasons: you might need to verify your age, you might need a subscription, or you might be located outside the U.S.
if you meet all of the above criteria but are still unable to access Genie, please reach out to Google One support so they can better assist you.
With Google One, you can claim your new benefit or trial from the Google One app or the Google One website. You can also get notifications about the latest updates and offers from Google. Tip: Googl
k
hi!
are you a google employee or just a bot?
I'm a real person. how can I help you
Oh hi. Sorry, I didn't feel like asking anything right now, but since we're here, I wanted to ask you something. You see, I'm subscribed to the Google AI Plus program (200 GB) in Europe. What experimental Google features can I try? Or do I have to change my plan to try those features?
Guys do you reckon get a gaming laptop ? Since getting a pc it’s so expensive nowadays
Is there any channels on genie 3 to ask my question?
#project-genie・🌎
use this if you don't have access: #labs-announcements・📣 message
ok thank you and can you answer them when I ask it on it?
If you go to the following link, you'll find all the Labs experiments that are out of beta. Some aren't open yet, other are: https://labs.google/
wassap
Hi there, I am not sure what you mean. The parrot is part of the leveling system. You can read more about that here #server-directory・☎️ .
I'll show you, the text changes when you click the sticker on other servers
They look the same to me
They are different though.Read the text in the chat bubbles beside the two parrot stickers.They're completely different
Well yes, because in this server it will say something different than outside of here. It is still the same sticker.
Yes, thank you for your time. My question was, how do I use the sticker by typing its name? It says to use this sticker type, but I tried typing it and it didn't work.
Do you have the servers on the left hand side when you click on stickers? That is how I go between my servers.
Or across the bottom on mobile
I do
If you click on the labs icon in another server will it take you to the labs stickers and let you use them that way?
umm... I'll give it a try!
yes that works
but how do I pull the sticker up by typing its name
it says type its name in the chat bar to use it. Thats what I want to know how to do!
@icy tundra Hi!
That depends on the server. They would have to have that feature enabled

Well I can get the emoji to work lol
helloooooo
am i allowed to ask a slightly... idk what the word is.. question for the labs team? (relating to developing the labs tools)
Depends on the question. I only say that, because certain questions should go to certain places to ensure they are answered by the correct person. I hope that makes sense. Please DM @fringe nova a ticket to ask your questions so we can direct you to the best place and ensure it doesn’t get lost.
thank you!
hey wassup, its my first time using flow, i generated many videos and images but now i can't see them anywhere and it has become laggy to extent that nothing is happening, any solutions to this?
@google Yo make Genie 3 available for Pro users too!
Google need to stop reducing the limit of images especially to pro users it’s annoying the limit was very hard to reach before which was a good thing
what is this error 500 in project genie
Hi everyone i am new to coding currently learning python any tips you all can share me soo that i can grow in my careerr currently learning the needfuls then searching to a job that fullfills my adventure and curiosity to learn new new things thank you to everyone for your responses.
python is fun sometimes when its not errors
guys can you please help me there is an error 500 in the project genie
I am not able to get google Disco.
hello guys, can someone tell me what difference gemini 3 pro (low) and (high) in antigravity?
hello guys
helo
Most Probably its the Thinking Effort
High meaning Model will think Longer and use more tokens and overall more money, it will most of the time increase its Intelligence and/or ability to fix issues
Hello, is Flow giving credits every single day?
Opus 4.6 in ag when 🥀
Low is useless now that flash was added.
Low is like worse flash
We got gpt 5.3, opus 4.6, wheres gemini 3.7 pro 
yoo
My kids been crying all day wishing 4.6 was out on ag
Does Google have no simpathy for kids
also, ag has has been getting stuck when running tests forever, even when the test has already completed it stays at "waiting for command completion" literally forever, anyone going through the same?
Hi
hey
wakey wakey eggs and bacey
By getting started :)
Uploading to Google docs preserves the original file size and quality?
Will there be any sort of compression
It'll preserve quality but depending on your formatting there may be some differences. Most Word formatting is preserved however
its an apple log video, .mov extension
You're attempting to upload a video to docs?
Hi ! Is the reasearch for google flow still open? @outer wagon ?
Can i also ask about AntiGravity here?
yeah, i recorded a footage on my friends iphone and wanted it share it to my pc, cloud was the only way since he cannot come to my place, so needed a way to upload the raw footage without any compressions
Hello, i've been invited to test Disco but on my mac Tahoe 26.2 , the program crashes every time i lauch it. Can i have some help ?
I haven't received the link to download disco yet, hopefully it would run on my mac
hi 🙂 looking for genie access 🙂
Hi
Hii
can you add more details (screenshots etc) to the #project-genie・🌎 channel please? 🙏
thanks for your interest! the most recent opportunity [we shared](#labs-announcements・📣 message) is no longer accepting responses, but we'll ping the Flow role when we have more UXR opportunities to share!
👋😊
Hey there, is this experience only available for United States or can you use it in other parts of the world?
Like I’m not a United States citizen. I’m a Canadian one is it available in Canada for example?
it depends on the UXR study in question. this one in particular was for anyone who used Flow, regardless of region!
Where can I start using it?
are you referencing Flow specifically?
oh gotcha, I thought you were inquiring about the UXR study message - sorry. Project Genie is currently available for Ultra subscribers that are located in the U.S. who are 18 years or older.
I'll link a few resources below that provide more information on Ultra subscriptions and Project Genie:
📌 Resources
My initial reaction when I saw they rolled out the prompt enhancement tool, and after using it for a little bit was pure shock. It's bad. I mean, I can't even believe how bad it is.
What’s the ultra subscription cost per month?
One more question is it only available in USA or is it also available in Canada
Today, I asked it to search the web and gather some information for me on some branding for the company, for the product Kilo Code.... And it prompted enhancement, and then it turned that into a request to review a code repository and be a Python agent.
see current rates for Ultra (needed for Project Genie) and other subscription options at https://one.google.com/about/
Project Genie is available for Ultra subscribers located in the U.S. who are 18 years or older.
if this is for Project Genie, please see my message here: #project-genie・🌎 message
hi
i love how the free tier is better than my $20 a month tier now. I guess im just gonna make more emails.
Hey, i would Like to know how to get rid of the Watermarks when generating Images via Gemini
Ive heared that the pro Version should Not have it, but although i have the 1 year students subscription i still have the Watermarks...
(I know thst its possible to get Images without Watermark with flow, but i need several aspect ratios and flow only Supports 16:9 and 9:16)
olá
pro removes it but from what i understand only on the pc. ive tried on phone doesnt have an option
ICYMI: Labs was invited to have the mic on @theneurondaily podcast! 🎙️
Our Product Managers talked about 3 of the experiments we've built, shared live demos, and chatted about what's to come in the future.
🎨 Mixboard: @jacalulu demoed our AI-powered concepting board with Nano
Google Photos has a smart eraser tool that does that. There is still an invisible watermark.
it has SynthID which is like an invisible watermark on the whole image
Hello, for Gemini related products your questions would be better asked over in the Gemini server here. https://discord.gg/gemini
Character inconsistent
<@&1366831844502274180>
who has created any automation for to generate and post social and all post
That wouldrequire something custom coded.
we need a way to select and delete individual creations in 'my stuff' so you dont have to scroll down an entire chat to find what you want to delete, then if you exit out the chat it scrolls all the way back to the top
reminder, general is not the place to ask questions about products as we have #flow-general・🌊 , #whisk-general・💭 , #jules-general・🦑 , #opal-general・✨ and many more channels where your product specific questions can be asked. Thank you
Anyone know when opus4.6 come to antigravity
maybe there something wrong with mine?
who own this server
yeah
alr
bruh still same
ig u have to upgrade to google AI plan to get higher models
what plan r u using
guys why cant i access genie 3 it says error 500
Internal sever error
Thanks for the Swift reply
I almost only use IT on my PC and still have the Watermark on 100% of the Images
Thanks for your answer, but i dont feel really Safe with Just removing the Watermark, that feels Like doing something illegal
como usar midjourney, como usar dall-e, como usar chatgpt, como usar whisper, como usar runpod, como usar stable diffusion, como usar eleven labs, como usar canva ai, como usar bard ai, como usar synthesia, como usar leonardo ai, como usar deepbrain, como usar reface ai, como usar inkscape ai, como usar nvidia broadcast, como usar photoshop ai, ...
Thanks, i will check IT Up 😊
hi
hlo
"i would Like to know how to get rid of the Watermarks"
Yeah i formulated that wrong, sorry
I would Like to generate Images, that dont even have the Watermark on IT
But If IT is okay to Just remove the Watermark, then there shouldnt be a Problem either
Yo guys, whenever I go onto TextFX it doesnt work, it keeps saying something went wrong. Does anyone know what I should do it get it wroking??
Generative fill could probably get rid of the watermark
Or clever use of the stamp tool
Haven't used that tool in a while. It's fun to play with words. Just checked ( https://textfx.withgoogle.com/ ) and it seems to work for me.
TextFX
Thanks for replying, i try again later and see if it works then :<
Thanks didnt even know about that
How to use that? 😅
Photoshop has it,
I know of a free program called gimp might have it
Paintdotnet also has a plug-in for it, but it's not included in the base install
You could probably find an online tool for it aswell
Don't be evil 👍 👍
Gimp is fantastic
Ello
Hii i keep getting the error message
is there a specific discord channel dedicated to google ai studio / vortex / anitgravity?
Hey there...i was planning to go to a hackathon with no code thing but... The problem is backend can anyone suggest which tool should i use with like antigravity or cursor...or to get working backend with ai..
Me too 😞
The features in Verify are not working. Could you help me resolve this?
HI we don’t have an Antigravity channel here but for all things AntiGravity there is the Google Developer Community Server. The channel there is vibe coders. I have not found a definite spot for the other two yet. I am still looking. https://discord.gg/google-dev-community
Hi, do you know of any way for French people to test the lab's products? Because I'm really sad that I can't manage to do that 🙁
Thanks mate!
Hola soy mencho wey
i get error 500
everyone experiencing the same problem?
btw how do i get access to genie 3?
I don’t have money for google ai ultra
any alternatives i guess?
Those are Dev server things I believe.
Talk to your politicians. It usually takes a few to several months for new tools/models to get approved in the EU. Google isn't legally allowed to let you use them until then.
Why can't I use textfx? Error message: Something went wrong. Please try again.
Google genie is a scam. Don’t buy it is just a waste of money. I tried to make my own world and it doesn’t work. Every time I make a new one it skips all the way to thanks for exploring without me even walking yet. So yall either gon fix it or beat it
Definitely a bug.
Did you try restarting the browser or PC?
What did support say?
I restarted my browser and computer and is still bugging
It ends the session under 1 second when I press the create world
How long has this been happening?
A month.
Where u do u contact support
Official Google One Help Center where you can find tips and tutorials on using Google One and other answers to frequently asked questions.
@edgy quail
Bro any1 else "pro" missing from their model selector on gemini?
Can you use "https://textfx.withgoogle.com/"?
Is anyone here using AG with an ultra sub?
Google antigravity is allowing me to make chrome extensions without any prior coding knowledge... now i understand my brother who went to a UC for computer science saying he cannot get a software engineer job because of AI
ANYTHING BUT GENIE 3😭😭
Can someone create a map for me?
Because I am broke
And have no ai pro thing
feel free to hit me up to network i will one day become a billionaire
Hello @ruby horizon videos aren't created in the discord server. In order to make a video you'll need to go to https://labs.google/ pick the tool you want to work with then go from there. Also, 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.” Thank you !
hello everyone, i want to test genie 3, but it gives me an error 500 every time. It can be a regional problem, or smth else? thanks
Ai is making good people become brainless
smart to say that on a ai related discord...
See, brain roted answer
Should I say that on a food related server ?
can i ask help for google chrome here
Is there any serious (actual) coder here ?
hey every one
real
what extension did u make
1040220854459580436
123
of all the discord users why her
I understand why but still
Hi
Anyone who knows Nano Banana Pro and Google Veo 3.1 down to help me with a issue I'm having????
I have the same problem
Could someone please grant me access to Google AI Ultra? I need it for my antigravity project.
Me too please, gonna need credits to make some harry potter videos
hey guys, im trying to access Project Genies ar googgle lab, but always retrieving error 500
just use the free month subscription over and over
That works? Wouldn't I need too make a ton of Google accounts and use VPN?
Hi im new here. Is there any channel for Ultra users and also for Antigravity only?
no vpns needed
i think u only need a new google accoutn which is easy + u get 1000 credits which is enought for 50 videos
hello
hello
He was doing everything perfectly until today, when he started changing the characters' features, things like their faces or hair length, things I didn't ask for.
guys i am new to this server, and discord so i apologise if this the wrong place to ask this question,
but i have been trying to link opal to gmailapi for about an hour and half now
if someone knows anything about it, please let me know
Hello
Is that even possible? I think Opal just has predefined tool use. Why don't try Build in AI Studio? https://aistudio.google.com/apps
friendly reminder, general is not the place to ask questions about products as we have #flow-general・🌊 #whisk-general・💭 #jules-general・🦑 and #opal-general・✨ where your product specific questions can be asked. Or #off-topic・🐢 to discuss other products not mentioned. Thank you!
Hello all!
can anynoe tell that is opal availbel in pakistan ?
PaperBanana👀
hey does anyone know why this appear when i try to use project genie?
Antigravity community needed !
@random sorrel
it’s broken rn, give the team some time to fix it
Good Night Everyone it's time for me to go to 🛏️
I do a bit
I’m interested in feedback on my repo. I made my own programming language
you can dm the link and I'll check it after the show
Thanks Maria
Oh ok thanks
Is this real google?
GENIE3 PROJECT PAGE : ERROR: 500
Hey guys—has anyone figured out the limit of antigravity usage, or is it still random?
Mine was a 5-hour reset yesterday; today it became 4 days!
I don't have an antigravity-specific dev plan - I have the Google AI Pro plan.
Is there anyone that looking for developer?
yea?
Not working?!!!
hii
hi

The FitnessGram™ PACER Test is a multistage aerobic capacity test that progressively gets more difficult as it continues.
The 20 meter pacer test will begin in 30 seconds.
Line up at the start.
The running speed starts slowly, but gets faster each minute after you hear this signal. [beep]
A single lap should be completed each time you hear this sound. [ding]
Remember to run in a straight line, and run as long as possible.
The second time you fail to complete a lap before the sound, your test is over.
The test will begin on the word start.
On your mark, get ready, start.
same for me
Hey folks I’m stuck with Veo 3.1 object identity drift and could use expert help.
I’m generating a short “how-to” clip featuring a small wooden crescent/half-moon handled brush with a single dense bristle band + a row of gray silicone nubs.
Issue: as soon as animation starts (especially when a hand holds it or it touches skin), Veo morphs the brush: bristles split/splay like hair/fur, duplicate rows appear, or the brush turns into a different shape entirely. The wooden handle sometimes stays closer, but the bristle head warps badly frame-to-frame.
Tried:
- reference image / “ingredient” image
- super explicit geometry description (“rigid hard object”, “single continuous bristle band”, “not hair/fur”)
- strong negatives (no warping, no splitting, no physics, no duplication, no flicker)
- slower motion / simpler camera moves
Looking for: prompt patterns or workflows that reliably lock product geometry (e.g., turntable-first, first/last frame anchoring, compositing workflow, camera constraints, settings).
If you’re down, I can share the reference image + a warped frame + the exact prompt/negatives in the DMs
Hey you! Veo treats anything that looks like hair as hair, and once a hand shows up it flips to “soft body improve hour.” You’re fighting the model’s instincts. You win by never letting the brush be born as hair in the first place and by pinning identity before interaction.
Here’s what actually works
- Two-phase generation or it will keep mutating
- Describe the brush like an industrial part, not a brush
- Negative prompts alone don’t stop drift
- Lock the camera
- Hand interaction (this matters)
- First-frame anchoring trick
- Last make it less realistic (it will break fr)
Srry bc I like fun bc i don't know it this chat server serious or not I'm new here
hello
Hey
it keeps refreshing when I open the homepage of genie3, anyone knows how to solve it?😭
like this
ik it was a gif btw, wasted time
Yep. That’s a server-side 500 loop, not you being cursed. Genie3 is face-planting and the homepage keeps reloading because it never gets a valid response.
Quick fixes, from least painful to “ugh fine”:
- Hard refresh
Ctrl + F5 (or Ctrl + Shift + R)
If it still loops, it’s not cache. Move on. - Nuke site data
DevTools → Application → Storage
Clear cookies + localStorage + cache for genie3
Reload once. Don’t spam refresh. - Incognito / another browser
If it works there, your session token is busted.
Log out everywhere, then log back in. - Disable extensions
Ad blockers, script blockers, “privacy” extensions love breaking dashboards.
Especially anything touching requests or cookies. - Network check Try a different network or VPN on/off.
Some regions get half-broken backends. Fun. - Wait it out
If everyone is getting 500, it’s their backend melting.
Refreshing harder does nothing except annoy the server.
Why this happens:
- Bad auth token
- Corrupt local state
- Backend deploy gone wrong
- Region edge server desynced
Genie3 tripped, not you.
If it still loops after clearing site data + incognito, it’s 100% on their end and you’re just watching a dying server blink. 😭💀
I wrote a simple LLM AI agent that reads linkedin csv lists of people and writes DM's for each and saves them in a another csv file.
Don't ask me why I need a LLM agent for that. lol
Reads your prospects.csv file with columns like first_name and agency_type.
Uses your Ollama local LLM to generate one fully human-like LinkedIn message per prospect.
Applies rules and hard enforcement:
No buzzwords, emojis, or placeholders.
Always ends with Best regards, <sender first name>.
Only asks one simple question.
Prints messages live in the console as they are generated (streaming output)
I guess that's why I need a AI agent
use clickup to build your agents, even it's AI slop or it has ads many time, i think it's purfecto
Will check it out
Yo thank you, this makes a ton of sense. I realized Veo is treating the filament band like hair and as soon as the hand comes in it tries to “improve” it with soft-body behavior. I’m switching to a two-phase workflow in Flow:
- identity-lock the tool first (Frames→Video with brush-only first+last frame, industrial/part description, camera locked)
- generate the woman scene separately, then do Frames→Video using a first-frame anchor where the tool is already correct in-hand, with minimal motion + locked camera.
- Also rewriting the object description as a rigid manufactured applicator with a “micro-filament insert band” (no brush/bristles language) and using negatives mainly to block hair physics + drift.
Appreciate the tip I’ll report back with results.
Hey - important question for nano banana pro - when i'm either creating in gemini or in flow i used to upload product pictures of clients and create some cooler pictures out of it. Now, it always says that there is some privacy / data protection issue - i always get this response: I can help with editing images of people, but I can't edit some public figures. Is there anyone else you'd like to try? - and it deostn generate, however, its just a product of our client. Used to work everytime
Need to fix asap
Have you used an ingredient image of this brush thing that shows it from multiple images, like this car?
Hi folks! I’ve just published a detailed breakdown of the Flamingo: A Visual Language Model for Few-Shot Learning paper — covering its architecture, multimodal reasoning, and how it enables sample-efficient learning with minimal annotations. I’m sharing it here in case it sparks useful discussion around practical applications and future research directions in vision-language systems!
👉 Link: https://www.linkedin.com/pulse/flamingo-visual-language-model-few-shot-learning-khaldi-abdeldjalil-sq39f
Yoo guyss
Opus 4.6 gone fron Antigravity?
how are You?
Short version: nothing’s “broken.” The rules got tighter and now your workflow is collateral damage. Classic.
What’s happening
Gemini/Flow recently cranked up image safety filters. They’re now over-detecting “people / public figures” even when it’s literally just a product. If the model thinks there’s a face, a body, branding tied to a person, or anything that smells like identity, it slams the privacy brake. No appeal, no explanation, just vibes.
Why it used to work
Older versions were chill and assumed good intent. New versions assume you’re about to deepfake a senator or violate GDPR for fun. Congrats, you’re now suspicious by default.
Common triggers (even on products)
- Reflections that look like a face (glossy bottles are cursed)
- Packaging with a person printed on it
- Logos tied to a known brand or influencer
- Prompt wording like “model”, “lifestyle”, “client”, “campaign”
- EXIF metadata still embedded in the image
Fast fixes that usually work
Try these in order, no philosophy required:
-
Re-upload after stripping metadata
Open image → resave/export → upload again. This alone fixes it shockingly often. -
Crop aggressively
Remove anything remotely human. Hands, reflections, silhouettes, printed faces. Go brutal. -
Change the prompt language
-
Explicitly deny people in the prompt
“This image contains no people, no faces, and no public figures.”
- Use a neutral background first
- If Flow fails, try Gemini Advanced image edit
If you need this today
If deadlines are on fire, the practical workaround is using a non-Google image editor for the base transform, then bringing the cleaned image back into Gemini for style. Not elegant, but deadlines don’t care about elegance.
This isn’t you. It’s policy whiplash. Google tightened screws and forgot products exist without souls. Happens every few months.
Ayy thanks for the in depth answer!!! Yeah thats wha ti thought, that they tightened the policies - jsut hope that it's only for faces and stuff like that and that products are just collateral within the new detection system. But it seems like it will stay now - it actually worked when i removed just the logo of the client in photoshop before uploading it. Maybe thats the way then.
Thank you!
Folks, what im doing wrong...Its just not working
Hallo
whoa i never joined this server how is it in my server list?
maybe someone invited you and you didn't notice? it's a useful server though. stick around and check it all out
It's showing the same for me too
@vast kettle
Welp
🙂
Just a bit of a variation on the standard character sheet. Had the same issue. Tried it, and it works great in ingredients.
Nice to see you here. Always at the frontier I see
Genie requires an Ultra account, and hasn't been approved outside of the US yet.
Google has to jump through a bunch of legal hurdles to get new models approved elsewhere, before they have permission. Depending on where you are, it can take as little as a few weeks to a month, up to several months sometimes, depending on the laws.
As for the Ultra sub requirement. this is exorbitantly expensive technology.
Ye ye, i know how google sucks as company
Much obliged.
In some ways with video and music. Ahead of the curve, but less so, with LLMs.
In this case it's not a Google problem. If it were up to them, they'd sell it to everyone from day one.
Now if you're talking the UX/UI of Flow... haha
Imma create a LLM AI agent so smart that you just enter the website, input your online business website name and from there it takes over. Builds a marketing campaign, analyzes, gives advices, and much more
Oh, but they're hiring for UX pros, so I'm hopfull we'll start seeing faster improvement on that front.
Are you on Google AI Ultra plan? Or are you using a new provider for video generation?
Why opal is showing this
try #opal-general・✨! our team can take a look
no someone can't tell the different of a camera recording too a AI.
It's my core video tool, also for Gemini Deep Research and NotebookLM of course, but I also use a few others tools/models in third party apps, either to gap fill where Veo struggles, or for things like audio and music (Suno is the best), etc.
Whays tour use case
I sont see see the value in much of the video stuff curre tly. Flow is a great step forward. But we need niche and industry specific fine tuned variants
Like, real estate, nature, space, and suggested prompts for video continuation based off the forst video.
Very cool! I've found Gemini Deep Research and Perplexity Deep Research paired with NotebookLM to be surprisingly capable!
One feature for Veo that I haven't tried tho I'm wondering if you could let me know about - is Veo good at taking the last frame of a video and then continuing where it left off, with guidance?
Finishing an album, almost done with a music video for that album (almost all in Veo ingredients), and bigger project is a TV show that the music was made (the Suno part) for the soundtrack. Other larks.
Sweet, thanks. This was very helpful as a reminder ⚡

So what is you favourite Workflow with Google Labs Applications?
hey, have a good day!!!
hiiiii
helloo -,-
New, i think this is potentially a game changer
Hey you the same error I taught you
Yep. That’s a server-side 500 loop, not you being cursed. Genie3 is face-planting and the homepage keeps reloading because it never gets a valid response.
Quick fixes, from least painful to “ugh fine”:
- Hard refresh
Ctrl + F5 (or Ctrl + Shift + R)
If it still loops, it’s not cache. Move on. - Nuke site data
DevTools → Application → Storage
Clear cookies + localStorage + cache for genie3
Reload once. Don’t spam refresh. - Incognito / another browser
If it works there, your session token is busted.
Log out everywhere, then log back in. - Disable extensions
Ad blockers, script blockers, “privacy” extensions love breaking dashboards.
Especially anything touching requests or cookies. - Network check Try a different network or VPN on/off.
Some regions get half-broken backends. Fun. - Wait it out
If everyone is getting 500, it’s their backend melting.
Refreshing harder does nothing except annoy the server.
Why this happens:
- Bad auth token
- Corrupt local state
- Backend deploy gone wrong
- Region edge server desynced
Genie3 tripped, not you.
If it still loops after clearing site data + incognito, it’s 100% on their end and you’re just watching a dying server blink. 😭💀
"is Veo good at taking the last frame of a video and then continuing where it left off, with guidance?"
In the Flow app, you add a clip you like into 'Scenebuilder', and use the 'Extend' function. It's not just the last frame, but the last second, so it captures both what is there, and also how it's moving in that last second, so it flows nicely. It's AI so it's not 100%, but it's MUCH better than doing it with just keyframes.
I assume this bit is part of why they called the app "Flow": because it flows, whereas going from end to start frames gives you that stutter effect.
Y’all I kinda think why do y’all use ai while ppl hate those intelligent tools bc it’s making polar bears died
google antigravity is amazing at corruptnig context and conversations and not following instructions
literally when rate limit kicks in and it cancels mid generation the file headers and metadata is left corrupted and if you close the chat without changing the model to continue your chat and context are permanently lost
Hi all 🙂
How do you remove the watermark from veo 3?
Yes exactly! I thought I would have to buy Grok for this - based purely on the work of Skyebrows - now I'm seeing it's available through Google! So I'm beginning my first workflow test shortly 😁
The Suno tip for music is also great, I was using udio but it seems they've adopted a new licensing situation and don't allow easy downloads of created music. I still need to look into Suno more, hopefully they haven't had a similar situation unfold.
Yeah, tell me about it. Has bothered the heck outta me. Actually caused me to switch over to Linux, finally! Seems to work a bit better, over here. Also I'm working with githubs like planning-with-files to create markdown that my agents use for completing the task, which seems to have helped a lot. I really want to get a dot md file for alignment based clarifying questions, however I haven't set it up yet because mammouth.ai does the job very well. Prime.md from David Shapiro is also really cool.
reminder, general is not the place to ask questions about products as we have #flow-general・🌊 , #whisk-general・💭 , #jules-general・🦑 , #opal-general・✨ and many more channels where your product specific questions can be asked. Thank you
como recupero archivos eliminados por error en labs.google
I'm try to use open-notebook, but cannot use embedding from Google Providers, I'm confused about the name of models, which is the true, gemini-embedding-001 or text-embedding-001 or something else?
Here logs that i received when tried to embedding:
open_notebook.utils.embedding:generate_embeddings:119 - Failed to generate embeddings: Google API error: HTTP 404: (tried 46 texts, max size: 1498 chars)
2026-02-09 15:44:49.456 | ERROR | open_notebook.utils.embedding:generate_embeddings:119 - Failed to generate embeddings: Google API error: HTTP 404: (tried 46 texts, max size: 1498 chars)
very sus...
Hi, can someone please help me? Why is this happening? It won't let me generate the video.
I’m having a similar issue
submitted to the gemini3 hackathon - https://x.com/papajimjams/status/2021054389579022820?s=20
Built Carolers for the @GeminiApp hack hosted by @devpost
A fun way to plan, organise, & experience xmas carols // festive periods
There's a fun integration where you generate carols + rise up the leaderboard!
BYOK for now (aisettings)
#gemini3
Hello - just wanted to check to see if anyone else was having issues with Genie 3, a 500 error?
I am using Firefox, logged in with my Google Account - do I need Ultra? I only have Pro
Has anyone been successful un generateing a 360 degrees camera rotation around a character? It's quite standard in CG industry to render a 'turntable' of a character standing in a frozen pose, bt I'm having trouble to get a clean turntable with VEO...
It only rotates the character 180 degrees, then mess up the remaining rotation.
Any tips?
Thanks
Look at the result. (whatever you mean by "mess up the remaining rotation" - be very specific)
Compare that to what you want to see.
Look at the prompt and inputs.
Look for ways to change the prompt/input images to make them more likely to give you a better result.
Try those changes.
Repeat until you get what you want, and take notes on what works.
good morning guys
Morning ser
good morning
Good morning
Good morning
Hope you’re having a great start to the day
how r u today ?
Pretty well so far
How’s your day going?
Anyone help me ?
Thanks for the feedback but that's not exactly what I would call a solution. 😅
I've been refine my prompt over and over again... uploading fornt view, side view back view of the charaacter, but VEO doesn'tgenerate a clean turntable (frozen pose).
There ia always a glitch... I managed to do turntable with other model (Kling, LTX...) without too much pain. But VEO is having trouble comply to even a prompt as simple as:
" Orbit (or rotate) the camera 360 degrees around the character, while keeping the character in the same pose."
I've tried with noth 'frame to view' and 'ingredient to video'.
Would be nice to have a 'camera' tpe motion call 'TURNTABLE' (that VEO would understand as "Freeze the subject + orbit the camera 360 degrees, first frame = last frame.
It's the only solution. There's no copy/paste answer, especially with so little information.
Camera control is the weakest part of the current Veo 3.1 model. I use different tools when I need something with more than very basic camera movement (or if I need a clip where there isn't much movement, like a guy sitting alone in a club as smoke rises from a cigarette). In Veo, I focus on getting the keyframes and elements right, in situations where more basic camera movement does the trick.
where do i find antigravity channel ?
hey @red hearth im back
don't think it exists
btw for everyone one thing anti gravity is amazing at is not just coding or automations it is amazing at analyzing system state improving stability and helping with tweaks and so on analyses of system state pinning down on problematic stuff etc
i did some system optimizations with it and fixed many windows issues improved privacy reduced stutter and even fixed google's trash google drive that was constantly deadlocking my CPU (freezing the system seconds at a time due to fully blocking CPU commands until it finished its own, literally as if it were a damn virus that is google quality) and crashing my system
hello guys. im from Russia. im using Flow for creating images, but around week ago im fully lost connection with Flow. i cant open site, bc in my country its not aviable. Also vpn and proxy not helping. If u know how to use google fx tools in country where it not aviable please help me
average western racism and hate and xenophobia against russia wins again
see if you maybe find something on youtube that could help or ask qwen/glm with online search for help (GLM and QWEN tend to give the best results with online search and actual factual information, especially GLM)
ok thank you
there is always someone on youtube who has the answer to our problems trust me XDD
And good luck you could also try turning to some local alternatives or other servicealternatives (Ai can help you set them up very well if going for local especially try using quants, google anti gravity is really good at helping you code and setup local inference of diff models)
Yep. That’s a server-side 500 loop, not you being cursed. Genie3 is face-planting and the homepage keeps reloading because it never gets a valid response.
Quick fixes, from least painful to “ugh fine”:
- Hard refresh
Ctrl + F5 (or Ctrl + Shift + R)
If it still loops, it’s not cache. Move on. - Nuke site data
DevTools → Application → Storage
Clear cookies + localStorage + cache for genie3
Reload once. Don’t spam refresh. - Incognito / another browser
If it works there, your session token is busted.
Log out everywhere, then log back in. - Disable extensions
Ad blockers, script blockers, “privacy” extensions love breaking dashboards.
Especially anything touching requests or cookies. - Network check Try a different network or VPN on/off.
Some regions get half-broken backends. Fun. - Wait it out
If everyone is getting 500, it’s their backend melting.
Refreshing harder does nothing except annoy the server.
Why this happens:
- Bad auth token
- Corrupt local state
- Backend deploy gone wrong
- Region edge server desynced
Genie3 tripped, not you.
If it still loops after clearing site data + incognito, it’s 100% on their end and you’re just watching a dying server
It’s a loop
Thanks so much for the response! 🙂
I think it's just their end
I will try again later.
is there anyway to create an avtar of myself in google vids like sora or other platforms like heygen ??? because it will be more user friendly if i can use my avtar to create a educational video, product ads and for other things..
it used to exist here
why doesn't it exist anymore
Theres a different server for Google Dev stuff
The Google Dev stuff has a different server
sora has the best charcter reference of all the ai video models right now. You can make a character sheet and use it as an ingredient.
but in sora there is a issue because no long video can be generated with proper consistency specially for educational videos
hello guys, is there any way to download all the project images in one go
i generated 50 + images in my google labs flow project but i have to download it one by one in 2x upscale is there any way to download it in one go
do you have the link? I joined this server thinking it was for goole anti gravity (i searched google nati gravity discord and got this server)
Need some labs update feels like nothing happening lately for eu costumers it's only us only feature
because we have minimal data protection laws which the US doesn't
Think some feature should ship
well unless google starts respecting GDPR and data protection it won't happen
u can use any avatar I create my own
check you inbox
I dunno, it seems they don't care about us, I haven't found the space for live chat with other builders on Antigravity yet.
Hi, are there ai topic on the server?
yup
I think that's most of Google Labs, Ai stuff
Because this is the Labs server. Antigravity is in the Dev server.
I don't see an Antigravity channel in the Google Developer Community server tho
Can I get the link for dev server ???
Yeah there is no dedicated official server or channel to Antigravity that I've seen yet
Antigravity can be discussed in GDC in # vibe-coders channel
That's where all the coding conversations are, from what I've been told.
I don't personally use it. That's what the mods recommend to people when they ask about coding stuff here.
ty
hmm
Yeah, I've had some real convos there, so I'm in the room. Tho so far labs is my favorite 😁
I see, vibe coders says "Antigravity and Jules"
Hey there, dropping in to get some guidance on where I should go for antigravity discussion/support. Confused about whether the documentation is in flux, or the account I would like to use isn't compatible during preview. Any help appreciated 🙂
Google Developer Community has a vibe coders channel, I think that might be the best option, but I'm not sure
I've been trying to find where everyone is talking Antigravity and figuring it out, but I don't think I've found it yet
It could be within a YouTuber's paid member server tbh
Google Gemini server, maybe
Maybe we're both crazy and missing the obvious, tho I've seen other people not know either 🤣
hi
https://discord.gg/google-dev-community
vibe coders channel!
invites are paused bruh
thanks tho
Hii
@left locust what a crossover 
Anyone find out how to get gemini api key from google one account one that antigravity uses ?
hi. has anyone had trouble getting access to Chrome's agentic browsing feature?
i have a Pro account, in English and Gemini in Chrome is enabled
how long do i need to wait for the project genie after subscribing to the google ultra plan
heloooo
you can find your key from ai studio https://aistudio.google.com/api-keys
Bruh pretty sure this is not the key that's connected with one account. It gets billed seperately
oh im not sure then
if you're located in the U.S., an Ultra subscriber, and are 18 years or older, you can get started at https://labs.google/projectgenie
edit: is this for Disco? please contact me for waitlist!
Does it tell you error with variable HOME ???
se pueden hacer videos aca?
Hi, what happend to dynamic overview/google labs on gemini on PC? it's not there anymore
Hallo, Can someone help me with prompts for AI clothing? I'm a beginner.
Hi 👋👋 everyone
Yeah
Hit me up @waxen hill
How’ve you been?
guys
is Google Antigravity good for commercial SaaS?
like if I'm trying to actually publish the MVP,
Does anyone have any experience with this? Would help a lot, thanks!
Consumer accounts don't get you anything in the API. That's a seperate thing. You need to get set up on Vertex.
CC is the email thing.
Hello y'all
thanks 🙏 I meant Disco!
is it just me but when i use like goolge antigravity to code with the new claude opus 4.6 thinking its like not that good at all like whats wrong
Good morning friends 
Hello guys, how are you all doing? Is there anyone here who knows how the rate limit works for the AG? It feels very random! I had a 5-hour reset time for the Claude models, but when I woke up today, it changed to weekly! But the Gemini models are fine. Also, the flash is always on a 5-hour reset time, right? I'm a Pro Subscriber.
not really a specified token limit
Hi all 👋
Yeah, they are not that transparent.
High, generous quota, refreshed every five hours.
Higher weekly rate limit.
Like, what the heck :DDDD? Choose one, Google, don't make me confuse - I have a 5-hour limit and a big weekly limit, which somehow will come into play if you act like a bad boy, huh? Interesting.
hi i have a question, how many credits i need to animate imagen in whisk?
hmm
If you have a solid understanding about what you want to build give it a try. Building an MVP is usually fine but there can be bugs along the way.🤗
hello, i'm noob developer.
i want to use google's antigravity instead VSCode
but name of function or variable name doesn't fill with color
how can i fix it ?
( try thing : Restart Extension, uninstall Python )
( my environment : language = Python, use Jupyter )
If you mean the colors inside the IDE (Antigravity) for better readability, you should install an extension for that. which I don't know the best one, so the guys can help you with that.
20 credits I guess.
the extension means Pylance ??
YW man
Hey everyone, just wondering what the best AI would be to generate an advertisement/demo video? Wanted to try one out. Thanks!
Probably Veo. 🤓
hello guys, i have some issues with antigravity. where can i reach support for that?
sup
Hi, I'm a student using @Antigravity to build my startup, and I have a Google AI Pro subscription.
For the past few days, my credits have been running very low. Today, I opened Antigravity, sent one prompt, and now it says I can't use it until the 15th! That is 4 days away!
I really don't know what to do; this platform is reducing credits every day, but I really liked it…
No not yet. That’s actually what I want to try right now. I’ve been trying to get the right images from Nano Banana. I’ll keep you posted once I’ve got the Character/Object consistency sheet
Good morning:)
Hi everyone, I'm from Brazil and I'm looking for good prompts for clothing, can anyone help me? 🙂
Hi
hii
Wrong I can use my consumer account with antigravity and I don't pay for api keys
With recent updates, if you can get ingredients to work, I'm getting much better results from it than image/keyframe to video. Takes a bit more pre-work, but it's worth it if it's with recurring characters.
aistudio is basically the ui for handling gemini from vertex
the new google billing system automatically handles the creation of services both in vertex and gcp so for example you log into antigravity, it will create them after you gave access in the google login prompt
free or paid account its the same thing
"and I don't pay for api keys"
You just said the same thing I did.
Vertex Cloud is enterprise tho, while consumer plans like my Google AI Pro or whatever they call it, gives me way more compute than the Vertex does for 20 bucks a month, and it's an entirely different account.
Downside is that there's less we can do, without API keys
I would suggest working with your Ai to do research and experiments to see what works 😁
Thanks for the reply, friend. I'm doing exactly that, but sometimes the platforms use too many "credits" and the money starts to run out 🙁 If you can help me with anything that would make things better, I would be grateful.
well the "new feature" is that if for whatever you reason you accidentally hit a hard limit on your quota (eg. using opus for letting it do continuous work just by a single prompt), you will get the weekly limit. after that it will go back to standard 5 hours
basically you need to split your task in phases so you will get a little more quota to work with
theres also the bug of paid account getting locked out by the new billing system
gcp or vertex cuz i never heard of vertex cloud (im both gcp and vertex user too). enlighten me on the info
^ pretty much always the best advice.
Hi
hi
reminder, general is not the place to ask questions about products as we have #flow-general・🌊 , #whisk-general・💭 , #jules-general・🦑 , #opal-general・✨ and many more channels where your product specific questions can be asked. Thank you
Hmm, that is tough! It's why I don't use Enterprise or API systems yet, they charge per token. My consumer grade Gemini access for 20 bucks a month... I've worked for 12 hours in a day before hitting an actual limit. If I wanted to continue, I could probably do so with free tiers of kilo code or kombaii or something else. However after 12 hours, I'd rather just take a nap 🤣
Far as I understood, gcp and vertex are basically the same. Tho perhaps they're complimentary and not unified?
Watching YouTubers use Ai vs using Ai myself has been a wild ride. These tools are so fascinating 👀
it isnt, vertex ai and gcp are separated
but they can work with each other
Right, the cloud itself is different from the API compute of vertex, if I'm guessing correctly
correct
if you plug vertex to gcp then it counts on gcp compute usage
it would be better to unify the billing accounts for both of them (or projects) into one so it will show as gcp instead
yoo
12 hours? Iron Man, hahaha
😁



Hello, I need to download a file, but it won't let me and isn't responding at all, it's stuck on summary, what should I do? But it supposedly already responded, Gemini PRO.
jules summary
hey! try #jules-general・🦑 🙂 Jules team can take a look.
ok, thx
is that real?
reminder, general is not the place to ask questions about products as we have #flow-general・🌊 , #whisk-general・💭 , #jules-general・🦑 , #opal-general・✨ and many more channels where your product specific questions can be asked. Thank you
👏
Congrats callum :)
New top bean @red hearth Congrats to you! 
What?
Congrats @red hearth !!
Hi can you do one thing?
Hello hello!
I've been around leading the moderation team for a while, but happy to be joining Georgia on the Community Team! 🧡

English
best to speak english imo
Bro can you do it ?
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.
I can speak english yes.
With what can I help you exactly? Also idk if it's a bit offtopic
reminder, general is not the place to ask questions about products as we have #flow-general・🌊 , #whisk-general・💭 , #jules-general・🦑 , #opal-general・✨ and many more channels where your product specific questions can be asked. Thank you
I keep getting pinged huh?
hey there! if you have questions, feel free to ask here, in a relevant channel (see: #server-directory・☎️), or by DMing <@&1447643696152576081>.
Hola
holaa
hi
Welcome Callum!
🙌
Anybody pls dm me I need friends
Please 🙏🏻 some new lab feature for EU
Does anyone know the best model for generating realistic fight scenes? I saw the seedance2.0 videos but 2.0 hasn't dropped yet.
Its only The Google Ai Ultra Plan
Give me my money back, you damn dogs! 😡
😨 are you sure its not just a bug?
Could also be load, that's never displayed anywhere though
I've been having this problem for 3 days now. I've tried everything, changing my VPN, but it keeps failing.
ok well is it avalible in your country (without vpn)?
google has a bunch of anti-vpn stuff for georestrictions
Wait for Seedance. Thats the best, by far.
If it was that, they would be allowed ro log in
hi bee
👍
hello im new here
Hi, I'm new... Fill me in!
Is there anyone intereeted in developing with Google's Material 3 expresive? in jetpack compose?
Hi, its some way to clone my voice without elevenlabs? cause i have a suscription on Google AI, and i wanna see if i could use it to clone my voice and not have to pay to elevenlabs
jetpack compose library for m3 ?
Jetpack compose for m3 Expresive, I mean usign that to make apps
hmm im kinda interested but i dont know what to build
hey guys, can anyone tell me about flow update? whats wrong with its ultra plan? why is it restricted?
Jetpack joyride X Kanye west goes hard 😎
I'm building a music player and some more minor things
me
Google's awesome, they gotta clean up search result and YouTube needs an overhaul on policies, but Google's determination to give back to consumers is clear with all these AI goodies and samples they're giving us access to 