#flow-images・🖼️
1 messages · Page 1 of 1 (latest)
Ayyoo
What is this place 😃 ?
with Whisk and ImageFX sunsetting into Flow, it's time for us to separate what was once the flow-gallery into two channels, #flow-videos・📹 and #flow-images・🖼️ 🙂
What about the sharing prompts/ingredients.
you're always welcome to share those too!
Even a collection from my website?
sure! if you make images and want to share the prompts, their ingredients, or something else related to that content, that works!
oohhh shiny
<@&1366831844502274180> This person keeps spamming prompts on every channel
please use #1238621511401734254 for any bug reports!
The comic...
I want Whisk back. Whisk wasn't actually transferred to Flow at all—it was completely deleted. Flow is hopeless. Why isn't the model that was on Whisk available there? That wasn't a transfer; it was a total removal. I want Whisk😢
Does anyone know a site similar to Whisk? Just not Flow. Flow is useless.
+1
but still trying to work well
No .
Lmao
I know. Dm me if you want
When I found out that Google was making a Flow app for Android, I had to download it, and this is my first picture through it.
And, to better reflect my interpretation of Google's announcement of the app, another picture of a studio, but it's under construction.
Who are these characters supposed to be? Lol
One of them is a Sonic OC, a mouse named Rachael "Roku-chan" Kiara/Yoshida. The other is a Pokémon/Sonic OC, a Jirachi named Lumi Stellara.
KATEGORI: Grafik
SAMMANFATTNING: Högst upplöst - utfört med teknisk precision: En biomekanisk hybridvarelse som förenar den....
TAGGAR: #NexusArchitect, #PureEssence, #PortfolioRedo
PROMPT:
Högst upplöst - utfört med teknisk precision: En biomekanisk hybridvarelse som förenar den mytiska silhuetten av en enhörning och en fenix genom avancerad Neural Style Transfer baserad på Giger-metoden. Genom att applicera arkitektonisk logik från en förtränad VGG-19-modell har de råa, organiska och mekaniska stil-egenskaperna från H.R. Giger extraherats och sömlöst integrerats i varelsens anatomi för maximal visuell koherens. Varelsen presenteras som en storslagen "Ur-fader" KARAKTÄR av syntetiskt liv, med en struktur bestående av ryggradsliknande rör, hydrauliska senor och ett exoskelett i polerad obsidian och benvit polymer. Enhörningens horn är transformerat till en fraktal, vass metallspiral med pulserande kablar, medan fenixens vingar består av överlappande, rakbladsvassa rör och genomskinliga membran som droppar av mörk viskositet. Ytan växlar mellan slemmig organisk mjukhet och kall industriell hårdhet. Belysningen är utförd i chiaroscuro, där djupa skuggor döljer komplexa mekaniska kammare och framhäver den tekniska texturen. Atmosfären är mörk, monokrom och auktoritär, präglad av en känsla av arkitektoniskt ursprung och teknologisk evolution. , makrofokus på biomekaniska kopplingar, rendering av syntetisk materia, komposition. Chiaroscuro-balans där högdagrar lyfter fram texturens mikroskopiska detaljer.
Prompt : A Liberty Walk 'Final Edition' Nissan GT-R in a darkened garage. The widebody is extreme, with exposed rivets and a massive swan-neck rear wing. Red rim-lighting traces the 'muscular hips' of the car. Cinematic smoke around the tires, industrial brutalist background, high contrast, sharp details on the Toyo Tires lettering, 8k resolution
@lavish oasis
High-speed automotive photography, a sleek silver Porsche 911 Turbo S speeding through a neon-lit Tokyo tunnel. Long exposure motion blur on the background and wheels while the car body remains crisp. Streaking tail lights, cinematic reflections on the paint, hyper-realistic, volumetric lighting, shot on Sony A7R IV, f/2.8.
Loving these
How am I supposed to find this?
Open up Google Canvas, tell it you wishes.
Thank you🙂
Fractal 😄
prompt : A night-time shot of a Mazda RX-7 FD with a Veilside Fortune widebody kit in Tokyo's Shibuya district. The car has a deep candy orange and black finish. High-contrast neon reflections from glowing signs dance across the exaggerated fenders. Rain-slicked asphalt, low-angle 'worm's eye' perspective, 35mm lens, cinematic grain
This car is very cool, On fire.....
I Agree
reminds me of Nina Williams from Tekken
McLaren 765LT, 1016 industries body kit, A dynamic Dutch angle shot, drifting through a rain-slicked Tokyo intersection at night. Neon signs reflect in the puddles. Heavy motion blur, aggressive camera tilt, cinematic lighting, 8k resolution, shot on 35mm lens.
I glad had LB bodykit
some one want participate in a work team for a mini series???? full AI project
i will need cars deftly
Go ahead make prompt
She looks a lot like Sarah Bryant of Virtua Fighter fame. Sick looking car there too! 😄
Lemme collab
prompt : Samus Aran in a blue bodysuit, stands near a Ferrari F40 with a Liberty Walk wide-body kit, while Sarah Bryant in a dark blue sleeveless bodysuit, stands near a Nissan GT-R R35 with a Veilside wide-body kit. They are in an upper parking lot—digital art.
Adding concept sketch design is fine
I suggest*
I’m just putting together a couple of designs. I need very futuristic concepts for the mini-series—high-end, next-generation super vehicles, since they’ll be the means of transportation for the ultra-wealthy.
Ok
And to truly feel like the future, it has to be highly innovative—because, for example, here in Dubai they’re already starting this year with drone-like flying taxis. So it needs to be at least ten steps ahead.
Ah i like plan make futuristic vehicles that awesome
proyecto_aeterna_2126 miniserie
Huh?
Where is she? Is she left?
I suppose to help
@brittle nimbus u can use edit image in flow mine use NB2 Pro
Before*
Edit : make hair realistic when I made
So I add Second reedit but very detail and realistic
I forgot PS in download image I use 2X as upscale
Hi Everyone, how can i use the ImageFX Model for generating images ? Because now it's everything on Google Flow right?
Imagefx & whisk are done and replaced by flow now flow can generate image after the two replace
but the results and the way to generate images are the same?
In flow after generate image same.
In flow there are choose select model nb2 pro, nb2, Imagen 4
Nice, thanks
Nice
Wow
[temporal logic: 8s duration.] [camera: ultra subtle your micro vibration. subtle dolly pull-back [instructions: enable z-depth persistence on subject hands, blade. ignore static anchors.] ACTION: [00:00] The subject instantly lifts the blade from their palm. and the hovering blackflame is extinguished immediately; and once the contact of the blade is released from the skin of the hand, the heat-waves around the sacred dagger begin to settle and dissipate. The subject’s fingertips, curl inward in a realistic, gesture of graceful, release and ease. PHYSICS: The heatwave distortion, gesture motion, and the sparking extinguish and collapse of the infernal black flame, these are fluid, and indistinguishable from lifelike physics
STYLE: Dark Sovereign aesthetic. masterpiece. 8k. cold clinical lilac lighting. deep, black shadowart cinema study. High-contrast foreground lighting. UHD skin textures (pores, micro-creases). No visible blood
https://labs.google/fx/tools/flow/shared/image/36c95726-5eff-4841-bee6-bb47ea9a320f
Hmm, I guess I did? Lets see if this one works.
Yeah, this one works, as long as it shows an embedded image on here, it means you got the right link 👍
Thanks for pointing that out, yeah, I should have noticed it failed to make a preview on here. Well, now you can enjoy my crude drawing of a castle! 😄
LOL I’m out of creative juice atm
I like it!
Can I use your prompt ? It’s very creative
Feel free! Spending time on Sora made me realize a creative person can probably reverse engineer any prompt they like with a few tries; so why even be secretive about them, I'd often just leave the prompt as the video title on there. I find it very interesting to see how other people prompt, it often leads me to trying new things or learning new terms!
I used to do that too! Sora got some amazing stuff there, God I miss sora already ,
Same here... Everyone's excited for that new Dreamina app which uses Seedance, but it's not out in the US and I'm too lazy to figure out how to get it on a VPN.
I also keep joking that, unlike OAI, Google is an actually profitable business... I suspect that in the long run Google will be a strong player in the AI space, including video gen, so I'm keeping an eye on Flow and some other projects here. I like the visual quality of Veo's output, but it's not up to Sora2's general utility level yet. Music and voices are bad, and the prompt interface needs more work, in my humble opinion.
I imported that to see what the input image was properly, that's a great result from that image, very cool! 😎
Wait a minute, can i use it to @frail halo?
Someone have a problem I can't download images 2k? i'm using gemini.google.com
Is there an error message that pops up? If so share it here and in the #1238621511401734254 so the team here can attend to it properly.
Also, if you switch browsers or a restart on your device may help.
Already did it but never worked at all
I would jot it down over in the #1238621511401734254, thats a usual fix but it may be bit more to it if thats the case.
Hi! We're so glad to have you here. To ensure our moderators can keep the community safe and everyone can join in the conversation, we ask that you please use English in the public channels. Thank you for your understanding!
No
does somebody know where i can reverse prompt a picture now, because i used to do it with Whisk but that got dropped ...
Gemini should be able to help you, or any other LLM.
Hi, may i knw if some of you can't generate 2K native images from nano banana?
Jajangmyeon Time 🍜
you can generate 4k quality images from nano banana 2 @leaden sapphire
You can try my app. 😅
https://g.co/gemini/share/81e4eef10879
Mansory! Alright that's some top end posh garage
Yup
Mother and Daughter 
do you have any other apps that simulate Whisk? btw, that tool you made is really handy.
Building Whisk in progress 😅...
wow
as soon as you have something, please let me know
it looks like you're doing a great job
Here's a Mobian Alcremie running her own bakery. Flow took some liberties (do Alcremie have tails?)
Mean while waiting, you can build a version of Whisk you want too. Just open Google Canvas, describe all the whisk functions you need, it would build a custom version for you .
A friend gave me this tip, and I actually followed it
It worked the day I did it, but after that it just stopped generating anything
Sometimes, adding to or editing the features can break the app
I even talked to my friend again; he mentioned the generation limit that was still in place, and I definitely hadn't exceeded it—plus, it had been about two days since I first used it
I opened another chat on Gemini and resubmitted a request through Canva, but the error still persists
There is a roll back function in the canvas , you can go back to the version was work eariler
The problem is that it was working fine before; it was only when I opened it the next day that it stopped working. I'll try something like that
thanks
Can you screenshot the errors?🤔
Maybe it would help others also run into this
for sure
I asked it to generate an image, and an error will pop up in a moment
"API request failed (Status 401). The reference image may be in an unsupported format, or the server may be overloaded."
English isn't my first language, which is why the language in the screenshot is different
Expired Session: Your login token or "handshake" with the server has timed out. This is the most common reason for a 401.
Simply refreshing the page or logging out and back in is the most effective way to "reset" an expired 401 token.
I've tried that
I've refreshed the page, cleared my cookies, and restarted my computer, but it just doesn't work
By the way, if you finish that Whisk project, please let me know 🤣
Do you have more than one Google account? You can forward it to a second account to test out.. 🤔
testing it right now
It didn't work; I'm still getting a 401 error
[17:39:16] [CONSOLE_ERROR] Error: HTTP Error: 401
at blob:https://4w0xtmt6lic7icchrk25byjgxn6y94pyl4olusmz1km421bh15-h906969431.scf.usercontent.goog/c8bc2062-8962-4ac6-be92-56dc104245db:1153:19
at Generator.next (<anonymous>)
at fulfilled (blob:https://4w0xtmt6lic7icchrk25byjgxn6y94pyl4olusmz1km421bh15-h906969431.scf.usercontent.goog/c8bc2062-8962-4ac6-be92-56dc104245db:913:26)
Go back to an earlier working version, 1 step at a time.
With some background code monitoring, it seems Nano Banana in canvas is acting up refusing to accept requests.... results 401 errors.
just not working
saving a prompt?
Hey everyone! Just a friendly reminder to keep all conversations in the supported language for this server. This helps our entire community engage effectively, and we've gone ahead and deleted the recent posts that didn't comply. Thanks for your cooperation!
What just happened here 🤣
😂
https://labs.google/fx/tools/flow/shared/image/cbc665bb-e24f-4dbd-8ebd-76c119a81989
Endless Trolly Problems!
Captain-Japan x Sonic the Hedgehog x Godzilla.
The ultimate adventure that crosses three universes, Cream, Sonic and Amy have learned that the Chaos Emeralds have somehow combined into one, however this strange event also brought two other worlds into Mobius, the worlds of Captain-Japan & Godzilla, destiny has chosen Cream to be the hero who would use the combined Emerald to save all three worlds with her friends Sonic and Amy, teaming up with Jamiko Miyamoto (aka Captain-Japan) to help her, while the Eggman-Empire and SHADOW have team up to use it for their own evil-plans, while the looming threat of Godzilla and the other Kaiju make things difficult for our heroes, Join them on the most epic adventure of all.
Prompt : A professional asphalt circuit at night under intense floodlights. High-speed motion blur, a glowing racing line on the tarmac, rain-slicked surface with vibrant reflections of neon track markers. Carbon fiber debris near the curbing, dramatic shadows, cinematic lighting, 8k resolution, shot on 35mm lens.
This is what I got on my end with the same prompt.
Some of the average works I made.
@woven cloudHey! Just a friendly reminder to keep all conversations in the supported language for this server. This helps our entire community engage effectively, and we've gone ahead and deleted the recent posts that didn't comply. Thanks for your cooperation!
Prompt : A narrow, winding Japanese mountain pass (Touge) converted for professional racing. Featuring bright yellow guardrails, "Slow" Kanji painted on the road, and thick cherry blossom petals littering the gutters. The perspective is a "chase-cam" view from behind a widebody car, focusing on the glowing red taillight streaks and the sheer drop into a misty valley. Photorealistic, 85mm lens, soft bokeh, vibrant nature.
Roving bombs.
@lavish oasis had to run 2 gens on that prompt for whatever reason the first one looked like the car was about to go over the ledge.
LMAO
Thunder McKing.
Here is another of my own characters that I've remade with Flow.
Invadoroid Robot K1 インバドーロイド is the final Invadoroid robot with missile-firing shoulders and both a pincer-arm and a sickle-arm for battle
Wrigley knows what I'm tawking about 🤘
https://labs.google/fx/tools/flow/shared/image/65b00010-9f2e-493b-a675-48b3e0e98f40
Prompt: A night-time shot red themed RGB 7k$ gaming setup from 3 foot vertical view. Dark style with triple monitor layout, cinematic shot.
I made a Zootopia apartment.
Prompt : Photorealistic motorsport photography of a Le Mans Hypercar racing (peugeot 9X8) through heavy rain at night on the Circuit de la Sarthe, water spraying from slick tires, dramatic track lighting reflecting off wet carbon fiber, subtle motion blur in background, shot on 50mm f/1.8 lens, cinematic contrast, editorial endurance racing style, ultra-detailed, professional trackside photography
Reinspired like this image but different scene and pose and character
what if split second cars in Street Version?
Trying something different, Kyodai heroes.
帮我生成一张蜘蛛侠海报照片
I'm sorry speak English this server use speaking English
what is bro saying
Oy, we don't speak mandarin mate
This is my take on what their saying:
- Help me generate a Spider-Man poster photo
or
- Generate a Spider-Man poster image
It's 2026 you'd think google translate would be infused better in this server. Their probably to busy raising more seed-money for the next AI model, more contracts, and/or w/e.
Either way it seems like their trying to generate an image of spiderman thinking posting that would either generate the image in server(Like the Arena server does). It seems like an immediate breach of copyright, I do know a bit of japanese but mandarin I fall face first.
Hope that helps. 🤘
Sorry, this is my first time using Discord. I thought that if I entered the prompt word in this window, the picture would be automatically generated.
No worries, thats the Arena server, and I get why you would expect it 😆.
labs.google/fx,Banana Pro in Flow seems to have no limit on the number of times it can be generated?
#rules-and-guidelines・📋 if this is your first time using discord read up on this channel to get up to date.
Thank you, I will
Prompt : Stunning Caucasian female warrior assassin, wearing revealing black leather outfit, lavender purple hair in top knot, confident pose, slim build, black leather bodysuit with cutouts and strappy, waist, hips details, dramatic lighting, cinematic composition, hyperrealistic render, tekken 8 fighting game character style, battle-ready stance.
I wish if I had tekken 8
Yup, the first gen spit Tekken out immediately. You reverse prompting over there or what?
Just Reedit prompt
Ahh, so its reedit, interesting.
Reedit*
Yup thats why i originally put Take 3..
Nice bro
Read up on the #rules-and-guidelines・📋 to understand how the server works.
Thx!
Oh @small crane
Hey @trim zealot, as Gaston mentioned, #rules-and-guidelines・📋 is a good way to start! As a friendly reminder we ask to keep all conversations in the supported language for this server. This helps our entire community engage effectively. We've gone ahead and deleted the recent messages that didn't comply.
Also please keep in mind that videos and images are not generated within the server, you'll need to head to https://labs.google/, pick a tool, and generate it there. Thank you!
prompt : A Subaru WRX STI rallycross car drifting through a wet, muddy corner under floodlights at night. Rain droplets freezing in mid-air, headlights cutting through the mist and spray. Reflections on the wet asphalt and puddles. Dark, moody atmosphere with neon sponsor logos glowing on the car body. High contrast, dramatic chiaroscuro lighting, photorealistic, Unreal Engine 5 render style, wide-angle perspective from track level.
Take 1 right here @lavish oasis:
The ultimate airplane.
It kind of has a B-2 spirit feel to it.
I had the idea of a giant super plane that you took a smaller one to get on and it would have way better seating.
So how many people does it seat then?
About 700 including crew, perfect for intercontinental flights.
Lmk we'll pitch it to UA together 🤘
hi
Nice
Uh all these use Google flow?
I loved on Korean comic or else
I made that one 1980's tv show cooler!
Normandy / Old Times
- My First Real Try with Gemini Pro
I Need prompt because I love it
Chat failed bro, could you refresh
Prompt : A Honda CRX Mightymouse Time Attack car attacking a corner at the Tsukuba Circuit. Glowing red brake rotors visible through the wheels. Intense "spitfire" exhaust flames downshifting. Aerodynamic vapor trails shedding off the rear wing. Motion blur on the wheels and background, sharp focus on the chassis. High contrast, cinematic lighting, gritty texture, wide-angle lens, sparks flying from the underbody hitting the curbing.
Puzzle version
Not bad ya, in flow should be Time attack spec
Have you played this yet @lavish oasis?
https://labs.google/fx/tools/flow/shared/image/cd5547ae-3b08-422e-84f8-e4395767516b
That why
The #1238621511401734254 is where you should express those concerns.
High-quality AI render of a Lego Technic box showcasing a Porsche 911 GT3 RS. The packaging highlights a detailed Lego model in Guards Red with a massive rear wing and carbon fiber texture bricks. Minimalist white gradient backdrop, official Lego logo, model number, and "Expert" builder tag. Studio lighting with subtle reflections, commercial product photography style, ultra-detailed.
no begging
A fine establishment.
The ultimate car, the Ultraknight.
Sorry English please I don't understand again
Ain't gonna kidding me, tell him make ai art prompt about this use Gemini or else
Prompt : Stunning female, assassin, wearing black assassin bodysuit outfit, lavender purple hair in buns, slim build, black assassin bodysuit with cutouts and strappy waist, torso details, snow setting, dramatic lighting, cinematic composition.
Hey everyone! Just a friendly reminder to keep all conversations in the supported language for this server. This helps our entire community engage effectively, and we've gone ahead and deleted the recent posts that didn't comply. Thanks for your cooperation!
Went with yellow & teal.
I like those side straps to her suit, and the hair color is dope AF
Shelbies are the best🤘
@lavish oasis she stuck the landing.
Its for compression, and the suit has a built in heat source.
Oddly the yellow + teal was a gamble that I also like here.
Prompt : A studio photography shot of a widebody BMW M4 Competition on a turntable. The car features a Mansory widebody kit, Swan-neck GT Wing, a vented hood, and yellow DRL headlights. Soft-box lighting highlights the aggressive curves and wide stance of the rear fenders. Clean white background, ray-traced reflections, ultra-realistic automotive rendering, 100mm macro lens.
I mistake Double word of widebody
Last one
Prompt : A widebody Porsche 911 (992) Turbo S (RWB) driving along a winding Amalfi Coast road during the golden hour. The car has flared wheel arches that blend seamlessly into the body, finished in a satin metallic "Frozen Berry" paint. Dust kicks up behind the tires, catching the sunlight. High-speed motion blur, realistic textures, volumetric lighting, shot with a DJI Mavic 3 Pro drone.
Please don't advertise your products here.
-
J'utilise seulement Gemini Pro
-
Ceci n'est pas de la publicité !
-
I use only Gemini Pro!
-
This is not an advertisement!
En anglais : "The bot translation is a hallucination. I never said I was selling anything. I am just a user showcasing Gemini Pro's amazing results. Stop the spam, Yaska!"
Traduction : La traduction du bot est une hallucination. Je n'ai jamais dit que je vendais quoi que ce soit. Je suis juste un utilisateur qui montre les résultats incroyables de Gemini Pro. Arrête ton spam, Yaska !
@Google Staff, please note that the bot Yaska is providing a completely wrong translation of my original message. I am NOT selling anything. I am just a French user showcasing the power of Gemini Pro and Veo. Vive la France ! 🇫🇷
.
Napoléon Bonaparte
- Not Perfect Yet?
- Practice makes perfect!
Gemini Pro / Android / Learning Outline...
Some 1950s vibe here
en plus c'est truffé d'erreurs historiques...
Did a series of "retro" images for the Mods and Rockers vids I did a couple of months ago
Quite enjoyable series of them to do
a girl
Make in Google labs flow not begging
a girl
Again please not begging
In The Stars
- If you learn, you will be
- May the Force be with you !
- Find strength you must !
Idk we may need a third opinion on this one.
yeah - like English is not my first language but I'm doing my best and how do I find tips and tricks
Prompt : BMW m2 G87 (LBWK Super silhouette) and BMW 320 Turbo Group 5, Drifting, with Cinematic Effects, 4K
Prompt: Mazda RX7 (panspeed), and 4 cars (2023 Honda CR-V (Time attack spec) , Toyota GR Supra (Rocket Bunny body Kit), 2024 BMW M2 G87 (Lbwk bodyKit), Nissan skyline gtr R32 (Auto Gallery Bodykit)). car stop
prompt : 2023 BMW M4 and 2023 BMW Z4, different body kit spec (tuners with widebody and Drift spec with GT spoiler), in Toys Figure Hot Wheels separate package, different cover package near different cars and different package
prompt: Nissan GTR R35, Nissan Silvia S15, Nissan 400Z, with different body kit specs (Veilside Body kit, Rocket Bunny Pandem Kit, Blitz Body Kit), in Takara Tomy die-cast, separate die-cast
We are the Mods
Flow is definitely getting better.
- What is his age today?
- Just an Illusion !
- Can an English speaker read French by default ?
- And an American ?
They really need a translator bot in here
No problem, I've accidentally done that several times lol
To be honest, I'm really struggling with Discord, and even more with English! 😅
- Oh tree YES for the Traductor !!!
@main creek my server has a translator if you'd like to join. you wouldn't be the only French person either
Prompt : Mazda RX7 with 2023 Ford Mustang GT, Corvette C7, Nissan Skyline GTR r34, 2024 BMW m4. each cars use different ultra widebody kit such as liberty walk, Pandem, Varis, Voltex. in hotwheels figure, background in parking lot
Ok, what kind of server is it ?
It will be nice for me..
It's called Dirty Dozens. 50 ish people from all over the world. Some that you saw in the first animated picture I sent you.
I already sent you the invite
Ok, Thanks (Merci)
here's a fun prompt for ya:
https://labs.google/fx/tools/flow/shared/image/49c8b7e1-9da1-4f27-adbc-67cc660ee46d
The most handsome man in the world.
Prompt : A candid Polaroid of a female mechanic standing in front of a modified Mazda RX-7 in a gritty industrial warehouse. The colors are desaturated and moody. The car has visible intercoolers and decals. Low-fidelity aesthetic, messy hair, oily aesthetic, authentic instant-film chemistry streaks.
Coffee Break !
Hey everyone! Just a friendly reminder to keep all conversations in the supported language for this server. For images and videos please include English subtitles or translations when posting. This helps our entire community engage effectively, and we've gone ahead and deleted the recent posts that didn't comply. Thanks for your cooperation!
A compilation of ideas I got.
@copper oxide Hey everyone! Just a friendly reminder to keep all conversations in the supported language for this server. This helps our entire community engage effectively, and we've gone ahead and deleted the recent posts that didn't comply. Thanks for your cooperation!
That "Rigitoni Squadron" one made me smile. All of these are very nice!
Let's just say it originated from a dream I had.
Your video inspired me to make these. Not sure if the world needed them, but here they are!
Well I'll be, Thomas looks accurate for the main part.
I did not specifically ask for him, but it crossed my mind as I was writing my prompt that there was a good chance it would key in on Thomas, just from describing where I wanted the face on it. Also, I said it was a blue train, that probably gave it a nudge.
I generated 4 there tho, and 2 came out not like Thomas at all, which is also kinda interesting.
Sweet!
New engine I made.
Hi @raven fable ! Just a friendly reminder to keep all conversations in the supported language for this server. This helps our entire community engage effectively, and we've gone ahead and deleted the recent posts that didn't comply. Thanks for your cooperation!
It's only possible in English, not in any other language 
The ultimate manga series where each volume has a different protagonist and world, but they all got there the same way.
prompt : 1980s Dakar rally classic (Peugeot 205 T16 style) navigating deep desert ruts. Sand-caked bodywork, vintage sponsor decals, raised suspension mid-compression. Warm analog film grain, retro motorsport illustration with bold shadows and sun-bleached palette, painterly realism.
Take 1:
Homemade transformer.
It must use English languange, honestly
My Glaceon OC, Glacier, is a multiverse traveller, so naturally, she has experience crossing between worlds. This is her in the world of Puella Magi Madoka Magica, shaking hands with Madoka Kaname, the anime's main character.
Puella Magi Madoka Magica property of Magica Quartet and Shaft
Glacier's seen a girl with a bow and arrow, and now she goes back in time (if you could call it that) to learn how to use one. She's in the world of InuYasha, and who better than Kikyo's reincarnation, Kagome, to give our resident Glaceon a few pointers.
InuYasha property of Rumiko Takahashi
Glacier's jumped into the world of My Little Pony Friendship Is Magic, and she has an intelligent discussion with Ponyville's resident alicorn, Twilight. But Glacier? "What are these 'elements of harmony'? Are they important?"
MLP:FiM property of Warner Bros. Discovery and Hasbro Studios
HELP! Hi it says my images transferred but under flow I only see videos. If I change it to imagefx it takes me to a welcome to flow page.
Also if I keep scrollig down the library page crashes
and everything is taking forever to load.
Error number 5

Quick fixes & shortcuts:
Free up RAM: Close other tabs.
Clear cache: Ctrl + Shift + Delete (Windows) or Cmd + Shift + Delete (Mac).
Incognito mode: Ctrl + Shift + N (Chrome/Edge) or Ctrl + Shift + P (Firefox).
Disable hardware acceleration: Settings > System.
Fun input!
He was quite a evil ball.
Better fix: take a few hours of your life to research and install Nobara (what I use), Bazzite, MintLinux, Ubuntu, or any other popular linux distro on your hardware, and reclaim like 30% ram just from being on a less bloated system. I switched 4+ mo ago, I'm not going back! Nobara and Bazzite support Steam as part of the install, Discord works, Brave browser is better privacy oriented while also supporting Chrome extensions, it's really not a problem these days! You can even ask your favorite LLM for help by sending it screenshots or text logs of what you're working on as you set it up, which imo makes things much easier these days! If that's too much to do, install Brave browser on windows and turn off annoying Windows features; you can ask a LLM how to do that also.
To keep this rant slightly Google focused: Google Antigrav supports Linux, and it's the main environment serious AI enthusiasts seem to be interested in!
None worked I still don’t see my images.
What are all these instructions for this to see my images inside of flow?
That’s all I need. I need to see my images. I can only see them when I’m creating a photo in nano banana.
And I use anti-gravity to fix a ton of stuff. This is not fixable by anti-gravity and I’m not installing extra stuff. I’m on Mac.
I fix computers so I want you to understand I’m not a regular user. I’m an advanced power user.
It’s in project titled imagefx in flow
I dont have that
I dont have it
u didnt even try look for it
are u gonna at least try do it right now
They are only in Nano Banana
are u gonna try it now or not
DUDE I WILL GIVE YOU A SCREEN RECORDING
screen recording of what
just to prove it
if imagefx creations appeared in ur flow assets list then it means the migration worked
"Imagefx Creations"
Because you didnt look in project assets, dont look in library page
its literally in home page of flow , not "My Library"
Image-Fx didnt have option to choose nano banana as model choice tho, thats only whisk.
are u confusing with whisk
sorry were those images were generated in image-fx tools or not?
Hold on...the video exceeds 100 mb
a screenshot is enough, or just use snipping tool to record video
windows key + shift + s
just look
i know I even show where I know it's at
what the hell
wait i realied u were speaking in the vid let me put on earphone
ah
so i assume u dont want it to be in the "asset list"
no that sucks
and u want it in the project right?
try this instead
in imagefx creations project page, do download project see if it works
wait no
that would just download , and not move to the project
so that's the thing a lot of my images were not inside of projects
I didn't use everything inside of view. I would just make photos to make photos.
for fun
main homepage https://labs.google/fx/tools/flow/
Flow is an AI filmmaking tool that lets you seamlessly create cinematic clips and scenes using Google DeepMind's most capable generative video model, Veo
For how far you have to scroll down to find the project folder titled “ImageFX Creations“ depends on how many projects you created, lol.
I dont have that
screen record pls
u have to look carefully
If it exists in your flow assets list, then the project folder page exists too.
Like i said , DONT LOOK IN "MY LIBRARY" page.
ok
LOOK in here https://labs.google/fx/tools/flow
Flow is an AI filmmaking tool that lets you seamlessly create cinematic clips and scenes using Google DeepMind's most capable generative video model, Veo
tbh, I don’t even know why they haven’t fixed the My Library page. It’s still super laggy and useless at this point.
LIFE SAVER thank you for persisting despite my stubborness lol
Phew I’m glad I was getting worried
Honestly, I blame whoever was in charge of this whole migration thing.
It’s messy. Also, I hope they add some kind of “pin project on top of the page” or something because damn.
Yeah that would help.
Something...anything but I was lost I didnt even notice it was there lol
Don’t worry, I also didn’t notice it was on the main page on my end.
Literally had to squint my eyes to look at the project title one by one.
I mean you have to have an eagle eye. I forgot what projects I have made over the year
I CAN RELATE TO THAT “FORGET WHICH PROJECT “
😭 I have bad habits of creating new projects all over again.
i maybe have made a projects with that SCP bottle...who knows? I make rando projects ALL THE TIME
The project titled as timestamp didn’t help at all too
LOL
I feel bad for being so stubborn so thank you. You must have some sort of like mother or father or brother or sister that might be stubborn just like me that's why you probably persisted. I don't think everybody would do that so I thank you for your patience and understanding.
You were totally right...
It honestly should be in the damn library
that's where they go
it was the first place I looked
Nah, it’s all good! I’m just glad you didn’t end up lashing out at me or something and give up on me like how other people here in this server did to me ahaha. I’m happy that your issue is resolved.
they let you add covers? That helps
yep its a newly added "quality of life" feature
its about time LOL
thank goodness
Well you are the MVP today! Im sure others will read our convo and learn something
I hope you have a great rest of your day or month or life....you deserve it.
Well, hopefully, the flow team will take notice of this and ... rectify or at least fix the My Library page or something.
That actually made my evening c: Thank you so much! I hope you had a wonderful day too.
I hope!
@wicked swallow u ok?
Fluffy shirt!
prompt : A sleek modern die-cast toy display rack holding 10 Hot Wheels-style cars, each meticulously modeled after real-life sports and luxury vehicles. The rack features clearly visible numbered compartments labeled 1 through 10, with each die-cast car neatly placed in its corresponding slot. Studio lighting highlights realistic metallic paint, glossy finishes, and intricate details like tiny wheels, grilles, and badges. Clean minimalist background, shallow depth of field, professional product photography style, ultra-detailed, 8k resolution.
prompt : A Porsche 911 GT3 RS with a wide-body silhouette is positioned inside a professional wind tunnel, illuminated by laser sheet lighting. Smooth, laminar airflow trails are rendered in glowing cyan and blue particle lines, flowing over the massive rear wing and underbody diffuser. The scene is captured with photorealistic, high-speed scientific photography, featuring cinematic rim lighting and presented in 4K resolution.
- Leonardo da Vinci
- 1452 / 1519
My OC using a handful of her powers
No beggings
Hello @modern relic no beggings, make on labs google flow
U seem wrong post
Flow is an AI filmmaking tool that lets you seamlessly create cinematic clips and scenes using Google DeepMind's most capable generative video model, Veo
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. Also images and content cannot be generated in this server. Head on over to https://labs.google/ pick a tool and start generating. Thank you!
A dignified and sharply intelligent Southeast Asian (Khmer) man in his late 30s. He possesses a strong, chiseled jawline with a well-groomed, short, dark scruffy beard and mustache. His skin is a warm, natural medium-bronze tone. He has striking, deep-set dark brown eyes that exude confidence, focus, and quiet analytical power. His eyebrows are thick and naturally arched. He has a distinct widow's peak hairline with short, neatly styled dark hair fading slightly at the sides. His expression is composed, serious yet approachable, carrying the aura of a visionary director or a master architect.
Locomotive I made.
Prompt : 1999 Nissan Skyline R34 GT-R race car drifting with a modern Nissan Skyline ER34 with an ultra-wide LBWK Super Silhouette body kit. Sharp angular fenders, exposed rivets, massive rear wings. Heavy tire smoke billowing from the rear wheels during a high-speed tandem drift on a wet Japanese mountain touge. Shot in cinematic high-contrast black and white monochrome photography, gritty texture, action sports shutter speed.
Prompt : Cinematic wide shot, make, standing alone in a quiet minimal park, back completely facing the camera, surrounded by soft green trees with dappled natural light filtering through leaves, no other people visible, calm and peaceful atmosphere, muted natural palette — soft greens, warm neutrals, shallow depth of field on background trees, completely still figure.
Two months ago Flow's NB2/Pro would listen to me perfectly like in here. A human's image is one thing, but to symmetrically get every little piece of armor from an android by prompt was just on another level.
To think it was about to hit even minimalistic prompt perfection before the recent "updates" and now it's about to be outperformed by GPT image 2. It still doesn't totally happen yet tho, so they still have time to emergency-backpedal/rollback the mistakes.
Average city setting.
prompt : Photorealistic 3/4 view of a Nissan Z (400Z) with a Liberty Walk (LBWK) Super Silhouette widebody kit, featuring a massive front air dam, extreme boxy fender flares, and a giant rear wing. The car is liveried in a classic red and black racing scheme with white sponsor decals. Shot in a brightly lit studio with a clean white floor, sharp reflections, high detail, cinematic lighting.
im guessing not everyone is having the quota Error code 253 thing then? cos people still making images?
My mood:
prompt : Choro Q style (super deformed, cute, stubby, short wheelbase) miniature versions of a Nissan Silvia S15 with an LBWK widebody kit and a Pagani Huayra BC. The two toy cars are drifting in tandem on a Japanese Touge mountain pass with sharp hairpins and cherry blossoms. Vibrant colors, playful toy-like appearance, high-quality 3D render style, dynamic action pose with smoke from the tires.
Use this car
Dacia logan?
Ye
challenge accept
Based on a dream.
prompt?
"1950's artwork of a housewife using a katana to fight Pokemon being sent in by a robber breaking into her house."
Well, Google Flow has added in a new feature: making your own characters!!! So, I've set up Glacier and Roku-chan, my two most recognized characters (on roleplay servers btw), and decided to try out the feature. Now, I can just click and drag the characters into the prompt.
This is Juno driving the Battle Bus from Fortnite. Originally this gen had guns and bullets in the picture, but after checking it against the rules of this server (see rule #6), I had every right to be concerned, so I had the AI replace all the weaponry with water guns and balloons.
Battle Bus gens and Fortnite related (don't expect much)
Made a book cover.
I'd buy.
turned my roblox creations into real stuff, its so fun
My Hero Mr Bean.
prompt : Choro Q style (super deformed,
with simple sketch LOL
Flow for image generation is so much fun.
It's especially useful for turning h andmade vector artwork into unique 3d items.
Hello, anyone know why I can’t use « reuse the prompt » when I click on a generated image ?
Things are getting updated
Ok
- But in France
- We have RGPD
Everything I made is from my tools I've created
- Very Good
The megazord is pondering hard.
- Knowing I will never be is my true wealth
Lol
Hello! You're lucky to have your own AI.
Gemini Pro helps me improve my ideas.
I’ve achieved my goal, and I’m testing
if it works across different fields
The feature will come back ?
Most likely yes
My own A I? It's just the tools I've created that's all within Flow
Yeah, sometimes automatic translations aren’t perfect—they can change the meaning!
Thanks
Hello guys, is there any easy way to generate images through nano banana without uploading manually? I used to work on other models with URL + prompt
Yes very different ways you can create yourself making an app under tools
Do you mean n8n app? But that will cost API 0.02 to 0.03 per image?
Not sure on the cost but inside of the Flow application there is a "Tool" feature that let's you code apps that can use NBP or something creating a use case for images etc
promopt : four best friends dancing in a bedroom, lead in baggy hoodie, VHS tape aesthetic, timestamp 02:14 AM, soft grain, warm lamp light, posters on wall, nostalgic candid
Not bad
Just For ???
How to provide reference images while creating images in flow,like there was option to add reference images in whisk???
hello, as a friendly reminder, videos and images are not generated within the server, you'll need to head to https://labs.google/, pick a tool, and generate it there. If you are asking for prompt advice please do so in the appropriate channel. Thank you!
Legion of Kansas superheroes.
Hes doin the cool slide
some sleek, unseelie, fae prince, images from the android app.
I'm sorry, but I don't have have access to /openclaw information.
Hello @midnight jungle make in google labs not begging post
prompt : Audi RS6 Avant in charcoal grey driving through rainy at night, street lights reflecting in puddles, moody cinematic car photography, shot from low angle, 35mm, f1.8, bokeh, water spray from wheels
wawa.
prompt : 2023 Ford Focus, rally spec, mid-air over gravel crest in Finland, dirt spray, action freeze frame, 300mm telephoto, overcast sky, photojournalistic rally photography
Hi, just trying the new update on FLOW, but do you know why he don’t create 2-3-4 images with one prompt ? I already check the 2 images in the parameters but he only generate 1
I just got this: is there a limit for Agent over 24h period?
"
You've reached your Tools Agent quota limit. Come back tomorrow or upgrade to chat more!"
If anything, the agent uses less compute I think than generation of images?
prompt : Cinematic low-angle tracking shot of a 2023 Toyota Crown Crossover rally-raid beast. Lifted stance, knobby tires, and mud-flecked bodywork tearing through sun-scorched dunes. Intense golden hour rim lighting, swirling dust clouds, and lens flares. Hyper-detailed, high-shutter speed clarity, capturing raw, high-octane desert racing energy.
prompt : Cinematic low-angle anamorphic tracking shot of a 2023 Toyota Crown Estate power-sliding through rain-slicked neon streets. Volumetric golden light catches swirling tire smoke and glistening metallic paint. High-speed motion blur, hyper-realistic asphalt textures, 8k resolution, dramatic atmospheric depth with sharp lens flares and vivid, wet reflections.
This one was solid, nice work @lavish oasis 👏
https://labs.google/fx/tools/flow/shared/image/60ea64fd-77ef-4ba4-bf0e-9ccc6d8d5747
prompt : Audi RS6 Avant in charcoal grey
Ty
May I repost
I use nb2 normal because this car has estate
What are you doing with FLOW ???? Google ??
I keep getting this error. Any suggestions?
Prompt: remove the background and place the kitten in space suit on Mars
Force refresh:
Press Ctrl + R.
This reloads interface.
Clear cache:
Press Ctrl + Shift + R.
This clears files.
Restart app:
Close Discord fully.
Launch it again.
Keyboard shortcuts will
likely have no
effect if the
server is down,
which is the
case in over
80% of situations.
System resource saturation
on an old computer
or a multitasking smartphone
causes these failures
in 80% of cases.
Something made from boredom.
Thread
Prompt : Cinematic double exposure: a Ferrari Purosangue fused with Shinjuku’s neon-drenched grid and Tuscany’s golden cypress hills. Ethereal layering blends cyberpunk cyan with amber Mediterranean light. Sharp 35mm lens clarity, metallic reflections, and soft-focus mist. An atmospheric collision of high-tech urban speed and rustic Italian elegance, hyper-detailed, 8k.
You got 3 hands 🤔..
I miss image fx, flow is very monotonous in terms of seed variation 
Helluva Hotel fanart.
Sorry yaska I erased I will fix
@swift halo
Looks normal now