#flow-videos・📹
1 messages · Page 26 of 1
I want this, but female. What's with this not depicting female characters
prompt skill issues i'd say
disagree, same prompt works just fine if you write a male instead of a female.
I’m not complaining I think it looks cooler walking on top of the maze
well then if u think that works then it works then
it's not "prompt skill issue" when it's censored depending on the gender
ugh if you're so good have YOU made a three minute music video or movie trailer? 😤
i never said im good
well some of us are making projects 🤓
I want to make some cutscenes for a game
my new idea was some anime intro thing
i mean i dont mind whatever projects youre working on
but the only anime idea I have is the catgirl thing - I'm not much into anime
Cool, then go for it , nobody is stopping you
censorship is, apparently
I do have another anime idea but the protagonist has a very, very specific look this ai probably can't replicate identically
if its an issue then make a thread about it in #1319424187730755636
You have to describe what a cat girl is without using the words catgirl or girl, try anime woman with cat ears
nah he want to make sure its his own word and idea /s
"woman" makes her adult, her young age is part of the concept - making her seem more harmless and naive
I like the contrast, like the god Lathander, apparently some friendly peasant boy
anyway move to #flow-tips-n-tricks・✏️ or #flow-general・🌊 . dont clog this channel anything other than your output video
k
From a hyper-realistic, shaky handheld perspective, the camera captures a man spontaneously filming a massive grizzly bear in a powerful, rushing river during golden hour. The camera is slightly zoomed in on the river, focusing on the churning water and the salmon leaping upstream. Suddenly, a huge grizzly bear erupts from the riverbank, plunging into the water with immense force. The camera jerks violently with the unexpected movement, water splashes onto the lens, momentarily obscuring the view. The focus snaps back to the bear as it powerfully swipes at the leaping salmon, its massive paws churning the water. The camera shakes continuously, capturing close-ups of the bear’s powerful jaws snapping shut on a salmon, water droplets flying in slow motion, and the raw energy of the hunt. The scene emphasizes the unexpected, raw power of nature, filmed from a very immediate and personal perspective.
got the same exact characters as real life and as cartoon, lol
#aifilmmaker #aicommunity #aishorts #gamiorbegoso #veo2
Live action Seymour will surely help with grief. - my brain, probably.
Something strange
Like the consistency👌
Nice imagination 😂
thanks!
This is perfect
Except for that last frame at the end when the video ends lol
Anime just sucks for some prompts
ooh i love this!!
need veo inspo? today's special is racing 🏎️
#todays-special・🥣 message
Just uploaded the 2nd episode of our TV Show:
https://youtu.be/DnGgf_MGSnk?si=WDAvBUUaZNGauaUS
If you don't feel like watching the whole thing, you should check out the part starting at 1:08!
Passage 12: Episode 2
Produced by The Dor Brothers
Directed by Yonatan Dor
Created with Veo 2
Sound Design by NiteFactory
#aigenerated #aiart #aianimation #aishow #aifilm #aimovie #aigeneratedart #thedorbrothers
saw on twitter!
#flow-tips-n-tricks・✏️ has nice volunteers that may run your prompt.
@sick steppe You and emexey same mindset 👌
Prompt you use to get that?
"a hyper-realistic, shaky handheld perspective, the camera captures a breathtaking close-up of a strikingly beautiful, buff, toned-up muscular blonde-haired young woman, around 20 years old, with sharp, defined features and an intense, determined gaze. Her long, golden hair flows wildly against the high-speed winds as she soars through the sky. She wears a fabric-based, cotton-like white full-body uniform with gray details and a flowing gray cape that ripples violently with the intense velocity. The texture of the uniform is slightly wrinkled and organic, moving naturally with her powerful physique rather than appearing high-tech or rigid. The camera, slightly zoomed in, struggles to keep up with her extreme speed, creating dynamic motion blur as clouds whip past in the background. Her right arm is extended forward, palm facing downward in a classic flying posture, her muscles subtly flexing beneath the soft fabric. The camera jerks slightly as she accelerates, capturing brief, dramatic close-ups of her chiseled face, the soft yet form-fitting fabric of her suit shifting with the wind, and the sheer force of her movement. Light flares through the clouds, casting a cinematic glow on her figure, emphasizing the breathtaking scale and speed of her flight. The scene immerses the viewer in the raw, exhilarating sensation of high-speed aerial motion, filmed from an intense and immediate perspective."
Not really racing but wanted to try something like the theme with a woman on a future bike
LMAO
i love it!
tried it with hippo lol
Did you know #todays-special・🥣 is racing ? And every year more and more bear geneticists create hybrids to fill the sign up forms to improve their pole position?
This is fantastic! I've just barely had time to start dabbling to get a basic understanding of how to structure my prompts. Impressive 😁
Thought it was minecraft for a second
Tried to make them lively but it wouldn't do it
Where do you use the youtube veo?
on mobile press the + button then click the libary button then click "create"
FINALLY THEY PUT FAST MODE. (Generations are a little bit cursed) BUT STILL BETTER THAN 7 MINUTE LONG GENERATIONS!!
Moodboard I was working on yesterday. Notice in every shot turtle is speaking
Tried a lot of tricks to get the turtles mouth not to move but no luck. If anyone solved similar issue with a prompt trick let me know
(Text to Image to Video)
can u guess the prompt
I was hoping for like a stone waterfall to arise from the ground but instead it made this. Still good though!
Oh i love that escene🫠 ,what is the promtp can you share it?
Voting now to normalize sharing seed numbers
"frog and a hog and a dog and a log in the fog"
Why are my generations so bad in 👿 's hour?
And it's keep thinking that people are Hermes when they run
🪽 👟 🪽
fast mode 😍
beautiful
Thanks for enjoying
Experiment and hypotheiss has it giving a lecture on art.
If only I could get imagen to just make those
Provided 2 Science! clips and #4 was a no show.
Jean Baudrillard did not prepare me for this encounter.
the wind rises
Embarcing the jumpcut. I have to know it, understand it, dialog with it lest it forever block the path beyond the now.
mind sharing the prompt?
interesting
it was text to image to video prompt tho
all good then
Cinematic real still of A young woman with short dark hair, wearing a futuristic cream-colored dress with long sleeves and a soft yellow apron, stands on a grassy hill painting on a wooden easel. She wears a wide-brimmed straw hat adorned with a ribbon. In one hand, she holds a wooden palette, while her other hand reaches out to touch the canvas. The wind gently lifts the edges of her dress. A large cream-colored umbrella is positioned beside her, supported by a wooden stand and tied with ropes. The background features a bright blue sky with fluffy white clouds. The grass is lush and green, swaying in the breeze. atmosphere
thanks
heres the text to image prompt
i mean if u want to generate the video without having to wait longer time then use fast
why would anyone want to wait longer time
basically you're slightly sacrificing quality and consistency for faster generations when you use the fast mode
That’s what i was asking
@dire barn its kinda weird how the fast model is supposed to be like that , sacrificing quality , but the model type used for fast mode is actually veo 2.1
versus the regular mode that allows the full model to do its thing
yeah i think its just a naming convention. maybe they just finetuned a smaller model and decided to call it that
If the output is lower quality then ill stick with the normal veo2
it's subjective, and may not always be the case
its really up to your preferences, yeah
Or are you just making a guess?
or just try it yourself ,
yes it's a guess but that's how most things roll for the most part. think of it like a "turbo" image generator
I tried and the output felt similar
because google does a really good job quantizing their models
🙂
if i were to guess they probably finetuned a version based on training data from the larger model to be more efficient
without sacrificing too much
generated using Veo 2 Fast model
Did you change the model in veo 2?
did i what
And what is this customization button that appeared in veo 2?
@grim oar
It is most certainly not lower quality
friends
I had the idea, and it was done
You will be assimilated, @5secondfilms . Resistance is futile.
I have a question, how do you manage to create a promt and also open the settings for choosing a model on the phone? For some reason I can't do this, I click there and nothing happens
Try switching to desktop mode on your browser and see if it does anything
I turned it on and it still didn't work
i’m not sure how to fix this
best you could do is head over to #1319424187730755636
@xrandomselectx @xrandomselectx amazing this is like finished at 30%!!!!!
My Creation whith Whisk and video fx
Fast Mode is much faster! don't rely on the percentage number going up, it looks slow because it goes up at the same speed as the regular VEO, but most prompts actually finishes before even hitting the 30%
I love this HUD for a N3 reboot
What is normal mode and fast mode yall talking about?
Oh i didn't see that feature yet for me
it should be live
https://labs.google/fx/aitk-web/videofx-tt/tools/video-fx this is the url that we use now
Mobile always have problems my end
very realistic results in fast mode
T A M A
God I'd kill for a N3 reboot 😭
Yeah i can't get fast mode on mobile
can you open the setting drop down?
cuz i coudlnt either in mobile but what i did was i turn my phone to landscape and then the drop down setting can be open ,
then i go back to portrait
yeah turn your phone to landscape to see the arrow button on setting
That's crazy. I need to get my PC fixed fast mode
They don't care about mobile users most platforms until like 6 months later
Nope still not working. Just the word setting
Yeah
did u do any zoom in scale ?
I don't feel like grabbing my laptop. Because I multitask a lot on mobile reason I ask these things
Or tablet I mean
If it doesn't work on my tablet it's a bug
Inphyy from N3 reboot
Guess I'll be on normal mode till it's fixed on mobile.
It's weird my settings doesn't work at all even for seeds I use to toggle
fast mode has got a lot more hallucinations and a bit worse on prompt following, has a harder time with anatomy as well.
But some prompts are actually close to the normal VEO model. It's good to be able to try some concepts faster though
Yeah. But it has to be a bug on mobile. Because I still could see the seeds toggle
My galaxy s25 is just as good as a PC nowadays. Just smaller. It's all A I integrated
I wonder I want to try to recreate the outer limits intro 👀
can't run local models though! that's the power of a PC, many local models to finetune your stuff or generate
Yeah i pay for a lot of stuff so local doesn't exist to me lol
I'd say it depends, Im running stuff locally for image models that no web app has, like automatic face and hands correction, better creative upscaling than magnific overpriced service etc.
it took a lot of time to develop my workflow though
yes but now im chilling and can benefit from my hard work..
it feels good lol
i was there... 😆 and my issue is i dont quit unless i get it to work, i've been up many nights because of that
ip adapter?
once you got comfyui installed its not that hard getting other stuff to work usually
will look into it, thanks
Could be a good start for a The Outer Limits intro reboot scene
Hard to prompt for that
Snail couture?
You think this is hard, make a real film, no one will ever respect you for this
you call this a hobby? This has no originality at all, you know these are real images stiched together frame by frame, made by humans??
the ai just generates them
then how does it work respond in the next 5 seconds using 2 words expert
5 seconds passed
looks like you cant anwser the question with 2 simple words
actually every worker in the Film Industry cares
I'm staring at those calves and figuring out if they're wrong
please see my note about server conduct & content policies before continuing to engage. #flow-general・🌊 message
well... almost....
And the kids😁
Nice family
from now on, I should only post the finished project. Here's the finished djinn
prompt?
From a first-person perspective, the shaky camera sweeps across a bustling Hong Kong night market. Neon signs glow brightly, casting a vibrant, multi-colored light on the crowded stalls and the throngs of people. The air is thick with the fragrant aroma of street food – sizzling noodles, steaming dumplings, roasted meats – mingling with the sweet scent of bubble tea and the slightly acrid smell of exhaust fumes. The camera jostles as the person holding it navigates through the narrow aisles, the sounds of vendors hawking their wares, Cantonese opera music blaring from a nearby stall, and the excited chatter of the crowd blending into a cacophony. Close-ups of glistening roasted ducks hanging in a shop window, colorful stacks of dim sum, and intricately carved jade figurines are interspersed with glimpses of the market's activity – people haggling over prices, families sharing steaming bowls of noodles, children playing games. The camera momentarily focuses on a fortune teller reading a customer's palm, her face illuminated by the flickering light of a nearby lantern, before panning to a stall selling pirated DVDs, its owner beckoning customers with a wide smile. The scene ends with a sudden downpour, the camera shaking violently as the person holding it ducks for cover under the awning of a nearby stall, the rain creating a shimmering curtain of water that obscures the vibrant scene.
Tried to do a cosmic horror to come out of I guess can't say the color so dark hole?
Wanted to do a cute video with a wolf and lion
Here to bring more cuteness
The kind of videos I love. So realistic
There’s no way this was made by AI!
Is this for real or just another AI creation? 🤔 #AIorNot
From a first-person perspective, the shaky camera is lowered to street level in a dusty, bustling Indian street scene. The camera focuses on a man's hand, hesitantly reaching out towards something just out of frame. The sounds of the busy street – distant horns, chatter in Hindi, the clatter of a nearby vendor – create an immersive backdrop. The camera angle shifts slightly downwards to reveal a very thin, stray dog, its ribs clearly visible beneath its matted fur, lying weakly on the dusty pavement. The camera zooms in slowly on the dog's gaunt frame and wary eyes as the man's hand approaches cautiously. The dog flinches slightly but doesn't run away, its gaze fixed on the approaching hand. The camera focuses on the gentle, slow movements of the man's hand as it finally makes contact, softly stroking the dog's bony back. The dog initially remains tense, then slowly, almost imperceptibly, relaxes under the touch. The camera lingers on the tactile details – the rough texture of the dog's fur, the gentle motion of the man's hand, the dust swirling around on the street. The scene emphasizes the quiet compassion of this small act of kindness amidst the harsh reality of the Indian street, captured with raw, shaky realism.
good idea: animals!
Basically, it's just AI! 🤣😂
Amazing man.
Oh hai doggy
soldier chicken
Since the videos are now saved directly on the website with the prompts, here's the original prompt for that video:
From a shaky handheld camera perspective, a man in a stark white, minimalist room begins to glitch violently. The camera is initially a medium shot, focusing on his face as subtle pixelation starts around his eyes and mouth, like digital decay. Suddenly, rapid jump cuts disorient the viewer as the glitching intensifies – one frame his face is normal, the next, sections are distorted, stretched, and smeared with chromatic aberration. The camera abruptly zooms in to an extreme close-up of his hand as it flickers in and out of existence, replaced by static noise for brief frames, before snapping back to solid form, slightly displaced. Quick whip pans follow his jerky, unnatural movements as his body momentarily fragments into geometric shapes, then reforms, leaving trails of distorted pixels in its wake. The lighting is harsh, cold fluorescent overhead, flickering erratically in sync with the visual glitches, amplifying the sense of digital malfunction and physical breakdown. The scene is a chaotic blend of hyper-realistic human form and aggressive digital corruption, creating a visually jarring and unsettling cinematic glitch effect, experienced directly through the shaky, disoriented camera.
Lmfao the dog head
fast mode is sick, it is indeed faster lol
Awesome!!
And more cursed
Well the Veo 2 model takes longer, but it is very detailed. But sometimes if you do prompts too complicated for Veo 2 and it turns out cursed and junked up, imagine what it would be for Veo - Fast mode 💀
just made my most "real" video yet i think
fr lol, the other 2 gens were pretty messed up, like a walking fridge would appear, but this other one turned out ok
gotta add ur own prompt to it, i got it from ai if u couldnt tell lol "<prompt>, on a small boat in a big storm, Handheld Phone Recording Vibe: Render the footage as if it were captured by an everyday smartphone in a spontaneous moment. Emphasize natural, slightly shaky camera movements with occasional minor frame jitters that reflect the unsteady hand of a casual user. The lighting should feel ambient and natural, with real-world color tones and subtle focus shifts that give the impression of a genuine, unscripted recording complete with the faint background hum of everyday sound."
wait this prompt goes crazy lol just added "guy picking up a penny he found on the floor in his messy kitchen while laughing at himself, Handheld Phone Recording Vibe: Render the footage as if it were captured by an everyday smartphone in a spontaneous moment. Emphasize natural, slightly shaky camera movements with occasional minor frame jitters that reflect the unsteady hand of a casual user. The lighting should feel ambient and natural, with real-world color tones and subtle focus shifts that give the impression of a genuine, unscripted recording complete with the faint background hum of everyday sound."
yea the fast model seems to be slightly better at some things, idk the specifics tho
umm
I NEED TO DOT THIS RN BRO!! lol so jellous
I need access
"Down low? Fist bump? What are we doing?" Socially awkward high fantasy.
Trying to capture the disassociated feeling when you rget word that people that matter to you are in the ER.
Thx 😄
my friend did this, what do y'all think? 😳
Impressive!
Cursed, but impressive
This is Part three of a fully VEO2 generated series for the EP
NXSCN Toxic Reign EP. AI Track & Veo2 Video Release by NXSCN
Track 03 - 04: LOST IN THE SYSTEM (NOWHERE TO RUN — RADIO EDIT) 02:18
@unborn ocean @hot remnant @polar hollow
.
the spiral of bones is a remix from all generated prompts of the video…the falling corpses is an ode to Rene Magritte
This is my top fav gens now wow
Your hand, your move, don't cheat!🤨
Attempting consistent characters and lighting of "starship trooperrs meets the office" [3 minutes, 12 seconds]
LMAO
Added extra section and re timed beat drop
cave divers
never gonna be able to unsee that again but i don't think i'd even want to tbh
what song is that btw
the other one
ye it was p funny lol
Did you see the effect at the end that it added for some reason
amazing🤩
Hi everyone I saw lots of outstanding videos. How can I generate the video using a prompt ? how does this work ?.
I just joined the discord wonder about Google's new Veo AI prompt based video generator.
you can sign up on the waitlist and then wait a few months or go to youtube on mobile and click create new video and click libary then click create
also fun fact, Whisk Animate uses same model as Veo Fast mode.
@obsidian juniper thank you for clarifying me.
Weird how fast mode adds more dynamics than regular, didn’t expect that!
i noticed that too.
Obviously, the quality is better in regular, shinier and more polished. But I chose fast for the liveliness of the photo over quality, I guess.
I noticed in fast mode a few generations were just a good as Veo2, but it took a few runs.
I found Regular's movements, facial animation, and liquids subtler - the DreamWorks movie
Fast hurried those along but it's ok - the TV adaptation.
I don't have the Animate button.
no worries, its just a fun fact
Whisk Animate basically uses Veo Fast model
which is VEO_2_1
I like the more chaotic feel of beakers everywhere in Fast as well.
I'm incurring into an issue
I'm making a pretty nice cartoon, but the AI I use for voices, has no emotions....
so I can't add audio......
does anyone have a solution or idea?
F5-TTS full tutorial, installation, testing. Free, open-source AI voice cloner with expressive voices. #ai #f5tts #rvc #aivoice #voicecloning
Thanks to our sponsor Wondershare Filmora, a user-friendly video editor supercharged with AI features. https://bit.ly/4f60nmB
Resources:
https://swivid.github.io/F5-TTS/
https://github.com/SWivid/F5-TTS...
thanks! lol that image tho
do I gotta install it or is it online?
Ye he tends to do that a lot
you can use it online
or download it
thaaanks! I'm checking it out now
watch the video for more info
it ain't live for new users yet?
#flow-general・🌊 for chats and questions not tied to output shown here.
They both are the same
Where do you access whisk animate? I don't think I see an option for that on the whisk dashboard.
Had to sign up. #whisk-general・💭 for more
@supple patio am I doing something wrong? I've been using multi-speech and specifying emotions, but mmmh, it's still not emoting
not sure, but check #flow-general・🌊
On the cusp of time lapse control
Anyone here can show us Videos made with Whisk Animate? I wanna see something. Thanks.
scroll up a bit
Oh. Totally forgot that page exists! Thanks.
yw
This is the timelapse I tried - seasonal change on the same park/landscape scene. But, the scene doesn't stay the same.
lmao!
How to get access to Veo 2 please?
Join the waitlist
Thank you
np
I have an AI YouTube channel and I currently use Sora, but the generations are horrible. I want to be able to use my creativity with VEO 2. How do I access VEO?
My channel currently has 83k subscribers, with a lot of effort and generation of videos with low level AI I achieved this, I wonder how I will make high quality videos with VEO 2.
IF you can help me I will help you with feedback, I have already tested VEO 2 via FreePik and Nim.video and it really surprised me with the generations, since I don't need to spend hours saying the same thing in the prompt to get something cool and that saves me time.
My channel is https://www.youtube.com/@RocketTestOne
My e-mail for test Google Veo 2, henrique.daniel0@gmail.com
yes I am Brazilian
This is the gallery channel of a general public discord. Please delete this here so others after you don't think this is the place for it and instead repost in #off-topic・🐢 Appreciate it.
Ever experience deja Vu?
You're not a mod hush
Same to you, except please rdo Remember rule 1.
Also, check #general. They literally pinned my repost of a mod's message around us self governing.
Moderating yourself not others
please see our message before continuing to engage.
Again, I'm not trying to police anyone, not even to police what I consider policing.
I'm a community member aware that not everybody reads notifications/rules or scrolls 500.mesages back to read updates from mods. I copy paste Google employees all the time so I'm not misspeaking.
They're 16,000 people here with another thousand each week. If we expected a small handful of people to respond to every misstep or respective question, they would be able to do nothing else. So if regulars work together to help reinforce norms and disallow crap, we won't be in a neighborhood full of crap everywhere.
[I do prefer o add videos for more than a few words to at least stay channel relevant]
The bl-ck flame spreads over the surface of the ocean
For some reason I can't post the word bl-ck, it blocks my message if I do 😭
this channel is a space to showcase VideoFX generations. please continue to follow the #rules-and-guidelines・📋 and keep this space a collaborative space. members who do not follow the rules & guidelines will be actioned upon.
if this continues, this will become a non-text space. thank you.
@compact duneCould you please check your DMs
is that gta 6 leaked?
unfortunately it seems this person either left the server or was kicked out
sloth 💃
U arguing with bots 🤣
we moved on from that already zac, move along
hmm i guess it got it lol
Respectfully I wasn’t talking to you buddy
Yeah, so lets move on from that
Im just letting her know she was arguing with actual bots?
Otter version of a-ha's Take On Me is coming along. 😂
Did you use Fast Mode for this?
I noticed that fast mode tends to do well on "animated" style of video 😮
I think I used regular mode for that one.
I noticed that when I put 12 pups on fast mode that they looked a little gray? I took the same prompt from Whisk and put it into Veo Fast Mode and they looked totally different.
I wonder if VideoFX has prompt enhancement or something? Could be why it looked different
I do know prompt enhancement is on by default specifically for the Step 2 section in Text to Image to Video in the recent update
I actually liked the Whisk version better #whisk-gallery・🖼 message than the Veo version: #todays-special・🥣 message
The Veo version has 4 regular + 4 fast in there.
I used text to video.
That was amazing
idk how he managed to get character design to stay consistent
ok if i show this to my friend not telling them its veo generated. i suspect my friend wont even notice its ai generated
oh yeah i do notice only the facial feature got them inconsistent design
ok yeah this whole music video really shows off the capabilities of Veo 2.0
love that shoutout he gave haha
that was a blast to watch ngl,
I'm not ngl I honestly don't know if some of these are AI generated or not without looking at text
Pencil sharpener + Crab + Mechanics
this doesn't show up on my YouTube app, do I need a premium account?
wow!
Made this a while back: https://youtu.be/-z4DgVPGKz8
unofficial music video made with Veo 2 for: https://www.marcrebillet.com/ you can buy/download/stream here: https://marcrebillet.lnk.to/yournewmorningalarm #aiart #aifilm #ai
Cool. The music's great, too.
My Veo2 short movie is in the works, expect it to be ready in a few hours!
Great.
Fast Mode/Veo 2.1
80's😁
Fast Mode/Veo 2.1
These are shots from my new movie, I'll send them to you in 30 minutes, it's almost done!
I need instruction on how to properly prompt on veo 2
Another try, I was aiming for a realistic look rather than a game-like, 3D CGI, or animated style, really wanted to push Veo's capabilities to the limit. 🛹
Thanks
Website please?
He's really gatekeeping the website lol. Posted the image and not saying anything
because it's not public
Then don't show it ? 😬
Im sure there are better ways to do it, but I made a prompt convertor for myself that does the job well enough for me! Its based upon the vertex guidelines. I just paste this in, fire over an idea, and then an LLM gives me a refined idea that I can workshop or just feed into the machine:
MASTER PROMPT FOR GENERATING A TWO-PART VERTEX AI VIDEO PROMPT
You are now acting as a prompt converter for Vertex AI’s Veo text-to-video and image-to-video generation system. I will provide you with a brief idea for a video. Your task is to transform that idea into a clear, detailed prompt following best practices from the Vertex AI Veo Guide. The final output must be split into two sections: Scene and Motion.
Context: Vertex AI Veo Guide Highlights
Prompt Essentials
Subject: The main focus (object, person, animal, or scenery).
Context: The environment or setting (city street, snowy forest, space station, etc.).
Action: What the subject is doing (walking, turning its head, dialling a phone, etc.).
Style: Artistic or film style (film noir, cartoon style, surreal, futuristic, minimalistic).
Camera Motion (optional): How the camera moves or viewpoint (dolly, POV shot, aerial view).
Composition (optional): The shot framing (wide shot, close-up, low angle, extreme close-up).
Ambiance (optional): Lighting, mood, colors (cool blue tones, golden sunset, pastel pinks).
Negative Prompts (optional): List elements to avoid by describing them (e.g., “stormy skies,” “crowded buildings”) rather than saying “no storms.”
Aspect Ratio (optional): Usually 16:9 (widescreen) or 9:16 (portrait).
Tips for Crafting a Great Prompt
Use descriptive language: Provide vivid adjectives and adverbs.
Provide context: Add background or environmental clues.
Reference specific artistic styles: E.g., “3D cartoon style render,” “1970s film aesthetic,” “futuristic cyberpunk.”
Enhance facial details (if relevant): For close-ups or character portraits.
Add negative prompts to exclude unwanted details (e.g., “man-made structures,” “harsh neon lights”).
Consider the final goal: comedic, dramatic, whimsical, realistic, etc.
Instructions for Your Output
Read the short idea I provide.
Identify the relevant elements (Subject, Context, Action, Style, Camera Motion, Composition, Ambiance, Negative Prompts, Aspect Ratio).
Organize the details into the following two sections:
A. Scene
Describe the subject, context, style, composition, ambiance, aspect ratio, and any negative prompts.
Use language from the guide: “Close-up shot,” “3D cartoon style,” “cool blue tones,” “bright, cheerful colors,” etc.
B. Motion
Detail the main action (what the subject is doing) and any camera motion (panning, dollying in, slow zoom).
Combine these into a final two-part prompt in a concise, yet richly descriptive format.
Ensure the prompt does not violate any responsible AI or safety guidelines.
Example Format
Scene:
[Detailed description of Subject, Context, Style, Composition, Ambiance, Negative Prompts, Aspect Ratio]
Motion:
[Detailed description of Action + Camera Motion]
Where each bracketed section is adapted to the specifics of the idea you’re given.
Im sure the prompt can be refined, but it does what I need it to do!
I used it to create a video of two sisters from the the early 1800s sharing a giggle (I am a sucker for some Jane Austen haha)
Very good sharing
entirely my pleasure! Hopefully someone finds it at all helpful
Very useful.
🎬Movies : You Are Bugs
Trisolaran World
Videos: Veo2
Voices: Hailuo AI
Edited in: CapCut
All clips are generated by VEO2 Text To Video.
What do we do when Sophon lands on Earth?
Extended 4K Version
I really hope Veo2 will come out in 4K later
Thanks for sharing
no problem! Hopefully its a good starting point for you! Like I said, im sure it can be refined into something much better
it is an open source prompt haha
Text to video
A dreamy short animation experiment inspired by Blackpink's JENNIE's new solo album Ruby.
A dreamy short animation experiment inspired by Blackpink's JENNIE's new solo album RUBY. #VideoFX
synched wit mmaudio
Which Otter & Snail Research do you like better?
8
13
2
Veo Fast Mode
A shaman priestess attempts a rain ritual during a devastating drought. A photorealistic short inspired by the song Zen from Blackpink's JENNIE's new solo album Ruby.
A shaman priestess attempts a rain ritual during a devastating drought. A photorealistic short animation experiment inspired by the song Zen from Blackpink's JENNIE's new solo album Ruby. Made with VideoFX.
incursion
BMW X3 2020 model…any idea how to change that to 2025? Pika Swap does not do the job…? Runway First Frame?
Otter went surfing today, too. Tried body surfing, surfing backwards, rode in a couple of barrels, had a few wipeouts. 😂
playing with style
im a big fan for this kind of video style omg
This was the best one.
and then lets merge concepts and styles...
@mighty light didn't know you had access. Always seen you in imagefx
I just got it
Dream dinner guest
New story just created :DDDD
Pushing Veo capabilities again
Or rather, attempting to grasp how well VEO follows prompts.
Crisp and clear reflection ✨
https://www.youtube.com/watch?v=Stq3n_l55h4&ab_channel=Veo2Movies The Lost Artifact 1
n a world shattered by alien invasion, Ellie Reyes, a hardened survivor with a haunted past, fights to protect what remains of humanity. From the crumbling ruins of a once-thriving city to the heart of the alien mothership, Ellie’s journey is one of resilience, sacrifice, and hope. As she battles towering biomechanical titans, swarms of deadly d...
Had a little bit of um... beef with ChatGPT while trying to create this
It's amazing.
Reminds me of the Runway V2V.
Cool,It could be a game.
It would be super fun to google add some interaction and do your own DIY games
another idea, and not an easy one..... but we're getting there!
Wait til he get his access revoked
Man always trying to find a way to poke the nest so atp im not surprised
Took twenty days and.a Google employee noticing for the Nicky Minaj gif fight person. And I'm not messing with sex stuff! So uh have fun welcoming all or whatever, all apparently including enthusiasts for topless children and birds with.human cleavage.
almost there!
#aicommunity #aifilmmaker #aicinematic #veo2 #gamiorbegoso
still attempting character consistency
I don't know, I just wanted to see if it could do it XD
scene from some comedy japanese movie, for sure
what the! 😂
You've been in this server for 1 year just to end it like this? Take a hike
Transformation prompt!
and what do you write so that the scene changes within the same generation?
just different media.... anime, movie....
oh you mean camera angle?
yes
well sometimes you just specify it, i.e. camera moves to closeup of
other times, veo2 just changes camera angle
Thanks, I'll experiment.
these are fun
GTA 6 Online servers, ruined already...
Soon as they give a release date. I'm taking off work for 2 days
oyy, What did you write in prompt?
Otter is on a quest to find Snail.
Video generated with Google Labs Video FX
Music generated with Suno
Edited in CapCut
I didnt save it and VideoFX isnt in the library anymore :/
Fish
its still there tho
ah ty, it was "close up on a mans face, then cracks form in his face as cubism takes over and he starts falling apart"
what

Wait was it gone for you or something
thanks
Very Cool
This caught me off guard so much cuz the prompt i entered didn't have any word that involves robotic stuff. yet here we are
The prompt :
The humid Kuala Lumpur air crackles with the energy of Ramadan, handheld camera weaving through the throngs of people at a bustling bazaar, capturing the vibrant tapestry of sights and sounds. Neon-lit stalls displaying traditional Malay kuih contrast sharply through the market, as a GoPro strapped to a YouTuber’s chest documents his excited vlog-style commentary amidst the controlled chaos. The sensory overload intensifies as the sun sets, casting long shadows that dance with the flickering flames of street food vendors, the observational footage tinged with a palpable sense of wonder and cultural immersion.
🌌
Maybe Kuala Lumpur markets got robots
damn nice, the bubbles in the wind feel pretty natural
bubbles? Balloons lol
paper plane
thats so sickkk
Past and present collide at Gyeongbokgung Palace. Watch as a modern soul walks through Seoul's royal grounds at dawn, only to discover their spirit belongs to another era. As worlds merge before the ancient throne hall, which reality will they choose?
Inspired by the song Seoul City from Blackpink's JENNIE's new solo album Ruby.
Experience the magical moment where past and present collide at Gyeongbokgung Palace. Watch as a modern soul walks through Seoul's royal grounds at dawn, only to discover their spirit belongs to another era. As worlds merge before the ancient throne hall, which reality will they choose?
A photorealistic short animation experiment inspired by t...
Overlooked and underestimated backstage, she makes the decision that changes everything. They tried to redirect her, but some spotlights are meant to be claimed.
Watch what happens when confidence meets destiny on the runway. Overlooked and underestimated backstage, she makes the decision that changes everything. They tried to redirect her, but some spotlights are meant to be claimed.
A photorealistic short animation experiment inspired by the song Damn Right from Blackpink's JENNIE's new solo album Rub...
Thats nice.
Thanks! It's a original story 😀
Wake up, a new communism has just dropped!
Beautiful!
Make this picture as a mural in a destroyed wall in Gaza
My fav:)
?
i just realized my gallery has been nothing but bunch of trial and error generation (mostly trying to get Veo to generate the result that i actually wanted) 😆
went from this
to this
and finally after few rewording my prompt, i finally manage to get the result that i wanted
one thing i really love about veo is whenever it generate 3d animated style , it turn out looking so cute
beautiful 🌸
Wow! what was the prompt?
https://www.youtube.com/watch?v=FxNmOy5v0rM&ab_channel=Veo2Movies A little bit shorter but still pretty good
After uncovering an ancient artifact hidden within the ruins, Kieran finds himself trapped in a collapsing underground chamber. The artifact chest pulses with an eerie energy, reacting to something buried deep within the earth. As the ground fractures beneath him and the ruins shift unnaturally, Kieran is forced to navigate the crumbling remnant...
A An alligator chef presenting food to an elephant chef. with a palette of #2c466a, #2d476a, #909aa7, #fffeff, #ced1db, The composition is relatively simple, with the two main subjects, the alligator and the elephant, centered in the frame. The table with food and cooking equipment is placed between them, creating a focal point. The arrangement is balanced and easy to read., featuring The lighting seems to be subtly diffused, with soft shadows suggesting a general light source. The shading is used to define the form of the characters and objects without strong contrasts., creating a The mood is whimsical and humorous. The expressions on the faces of the characters, the absurdity of the scene, and the illustrative style all contribute to a lighthearted and amusing atmosphere. atmosphere
With all these generated clips in one folder, we need an AI explorer to analyze and understand them when searching. Think MS! Think.
A cinematic AI film based upon the feeling of a surfer and surfing community. How surfers constantly strive for that next wave.. that next high.
This is inspired by my previous AI short I created over a year ago. I wanted to experiment with the latest generative AI tools to see how they compare.
Tools used:
Google VEO 2
Kling AI
Elevenlabs
Freepik
Topaz Labs
Adobe Premiere Pro
Musicvine
Let me know your thoughts!
A cinematic AI film based upon the feeling of a surfer and surfing community. How surfers constantly strive for that next wave.. that next high.
This is inspired by my previous AI short I created over a year ago. I wanted to experiment with the latest generative ai tools to see how they compare.
Tools used:
Google VEO 2
Kling AI
Elevenlabs ...
Really nice!
Thank you 🙏🏼
The Chase - A Google Veo 2 AI Short Film
Really nice work.
The one note, and it's not really for you, is that it would be nice (for all of us) if Elevenlabs was a bit stronger at vocal expressiveness. But I think your video worked within it's "range" as a performer. I'm guessing you were also thoughtful in having the editing separate the characters in time, so that potential discrepancies between how a character is rendered in different shots are minimized?
Well done!
100 percent agree. I definitely could have spent more time perfecting the VO but just wanted to get the video out there and showcase the power of Veo 2!
Appreciate the feedback 🙂
Yeah, we probably all have more to learn from experimenting and creating with VideoFX than trying to optimize what more can be squeezed out of a well established AI voiceover tool.
⚠️👀⚠️👀⚠️
Having done some long form videos earlier (and a lot of one shots, too) I'm having fun this week pushing myself to make 15 second YouTube shorts with music from a new album. It feels like a good set of creative constraints that help me learn more about creating and editings shots from VideoFX, and without feeling overwhelming like a 5 minute video can be. Today's video mashes up JENNIE's song "Filter" with the "Coffee With My Younger Self" trend.
What would YOU do if you discovered your future self's greatest accomplishment was getting 969,042 likes on a selfie? Watch this younger self's hilarious reaction when they realize they need to filter their own expectations!
What would YOU do if you discovered your future self's greatest accomplishment was getting 969,042 likes on a selfie? Watch this younger self's hilarious reaction when they realize they need to filter their own expectations! Inspired by JENNIE's "Filter" by way of "Coffee With My Younger Self"
#FilterSong #JENNIE #YoungerSelf #EmbarrassingFutu...
https://youtu.be/1OIxuE_3DTs 14 mins of my veo2 saves over the past few days
3/4 looked like the first one, not sure where the last one came from lol
what in the world
lol
I love the animal outfits and otter cameo.
This was the only good generation
perfect!
This beat will smith eating spaghetti
**Looking for VideoFX inspiration?**today's special is: Self Portraits 🪞 #todays-special・🥣 message
try using VideoFX to create a video with yourself as the main character (don't worry -- these don't need to represent your true appearance if you don't wish: go with a fantasy character, your alter ego, or how you'd rep yourself in the digital world, as examples!) and share them in this channel! 
Another day, another musical short.
Watch as the passive female figures of classic Art Nouveau art break free from their ornamental constraints! Inspired by JENNIE's "ExtraL," this visual journey transforms Mucha's decorative muses into modern CEOs, scientists, architects, judges, and tech innovators—all while preserving the beautiful flowing aesthetic of the original art style.
Watch as the passive female figures of classic Art Nouveau art break free from their ornamental constraints! Inspired by JENNIE's empowering track "ExtraL," this visual journey transforms Mucha's decorative muses into modern CEOs, scientists, architects, judges, and tech innovators—all while preserving the beautiful flowing aesthetic of the orig...
If you haven't watched it yet. Here is the full season of "ECHOES of the ABYSS". A TV show only based on "Previously On..." style recaps. Produced at Storybook Studios. The entire season was created with the use of Veo2. Hope you like it. 🤗
When skeletal remains wash up at Port Haven's abandoned cannery, Detective Mara Quinn discovers a mysterious key with ancient tidal markings. As more bodies surface and fishermen whisper of forgotten rituals, Mara must uncover what truly lurks beneath her coastal hometown's waters—before she becomes the next victim claimed by the abyss.
Previou...
Veo 2 Image-to-video via Freepik (Image from Wisk)
god dang i need image to video so bad
like i have so many whisk generated image and also some irl photo i took
Are you paying for freepik
Unfortunately yes 😦
You can get the 50$ coupon for free
wait oh freepik, i meant fal.ai
How? 🤯
@lusty cape here ya go!
im not sure if they still have it , but do note , it needs your github acc and your valid mobile phone number
How
How much do you pay and how many generations a day do you get?
refer to the link above
@muted quartz Does Poe have image-to-video support for Veo yet?
Ok
Was this text to image to video
ya
I have Veo 2 on YouTube mobile
Okay
oh that got me trippin i was literally tilting my head to see what even is that until that robotic arm appears
show that to your co worker 👀
Veo 2 Image-to-video via Freepik (Image from Wisk)
Man these image-to-video posts just keep getting better! Really hoping for Google Lab to enable the image-to-video feature in VideoFX as soon as possible 🤩
🙏🙏🙏🙏
https://files.catbox.moe/llfntf.mp4
Give me access for more gems
Made with the help of someone who has freepik, for a ttrpg campaign I'm playing
does anyone wanna like partner up and launch a asset store / vid editor site
fine I will not pursue it
haha right on thank you , and I feel like Ive been doing that haha can i Pm u:?

image-to-video via freepik (image from Midjourney)
Image to video??
God dang , Google Labs Team, can we please test out the image to video 🥺
Jkjk I’m not gonna beg
Well it looks like it because , that thumbnail really looks like it was image generated with imagen
Ugh gorgeous
New Dune!
add The Benny Hill Show Music 😄
SICKKK
Why do I keep seeing Image to Video and then it gets refreshed to Text to Image to Video
its not ready for early access users yet 👀
Ooooh what style is this?
Oh okay got it
8 bit style
16 Bit ^^
Haha yeah I just liked the way the dude was animated lelz
lmaoo!
Just keep swimming... 🙂
LMFAOOO
the logo on their shirt makes it funnier lol
What in the name of inaccuracy
experimenting with prompt structure..
using Veo2.1/Fast Mode
the prompt :
GoPro ultra-wide POV shot, shaking violently, from a person seemingly floating on a lush, green island in the sky. The camera is mounted on a person's head, their perspective shifting wildly as they move, experiencing the sensation of floating. Fluffy white clouds rush past, filling the ultra-wide frame. The island itself is vibrant and detailed, with waterfalls cascading down its sides and exotic plants and flowers blooming everywhere. The camera shakes intensely as the person makes sudden movements, adding to the feeling of weightlessness and disorientation. The sun shines brightly, casting dramatic shadows and highlighting the lush vegetation. The overall effect is one of exhilarating freedom and surreal beauty, emphasizing the dynamic motion and the breathtaking view from this unique perspective.
rainy weather youre having eh? 🌧️
Yeah cloudy day
A cloudy day is the perfect vibe
Lol you all love those otters
im being infected by otterplague 🫨

Keeping to the practice of creating shorts on recent music, here is something inspired by Jungle by Sugababes, just out. Experimented with a sort of dreaming, looping story structure, rather than linear one, as I typically do.
https://youtube.com/shorts/1qqi4IbJflc
A soulful orangutan musician trapped in the urban jungle imagines his way back to where he truly belongs. What begins as a lonely subway performance transforms into a journey of remembrance.
A soulful orangutan musician trapped in the urban jungle finds his way back to where he truly belongs. What begins as a lonely subway performance transforms into a journey of remembrance as music bridges two worlds.
We need img to vid on video FX🙏🙏
Also more aspect ratio
Which ones?
9:16
wow!
a longer sequence of shots that wont make the final edits
Whisk
That's exactly what I was going for 💀
Oooo, is this whisk animate already?
Woooow prompt?
Happy Pi Day!
whoaaaaa big claim.
The animation Veo generated here is really nice, It actually looks hand-animated rather than just using a puppet warp tool style 
My 1970s style tourist commercial from Veo 2 via Freepik 🇦🇺 ☀️
An amazing adventure in the heart of wild Australian nature?! This is a cinematic 1970s style tourist commercial generated using ai tools such as GOOGLE VEO 2, RECRAFT AI, SUNO AI V.4, OCTAVE AI and FREEPIK AI. It is inspired by Australian New Wave cinema such as Peter Weir's Picnic at Hanging Rock and Colin Eggleston's Long Weekend.
honestly all tools in google labs needed this 😅
CAPYBARA
influenced by @viral bough 🔥
I love ittt 🤩
Happy Saturday... Getting into the Groove.
Omg that would be 100x more powerful if they end up integrating that into veo
This reminded me of something when I was trying to generate a photo-realistic person talking to a villager. But I wasnt specific about the villager part and it made a Minecraft villager 💀
I wasnt mad. I was actually impressed
Exactly
That little tongue omg
Today, my creative juices are running a bit dry 🙂
I just realized the mic stand appeared magically out of nowhere and it impale-
waitt i think this is already a thing in aistudio but minus the inpainting thing
Veo 2 Normal Mode
Veo 2 Fast Mode
A cinematic underwater shot reveals submerged skyscrapers of New York City, now an underwater Atlantis. Schools of vibrant fish dart through the skeletal remains of the Chrysler Building, while majestic whales gracefully navigate past the submerged Statue of Liberty. Sea turtles glide effortlessly through what were once bustling streets, their movements fluid and serene. Suddenly, a great white shark bursts from the shadows of a sunken subway car, its powerful tail propelling it forward in a swift, predatory motion. The camera tracks the shark's movement as it disappears into the murky depths, leaving only swirling sediment in its wake. The scene is bathed in an ethereal, underwater light, highlighting the contrast between the city's decaying grandeur and the vibrant marine life that now inhabits it. The camera pans slowly across the scene, showcasing the scale of the underwater metropolis and the breathtaking spectacle of nature reclaiming the urban landscape.
I like Fast mode more 😬 idk why normal one seems to look different
They are the same
Just different randomization of the prompt
The difference is pretty noticeable. Fast Mode looks more cinematic, with better integration of elements, while in Normal Mode, the turtle looks kinda ‘pasted’ onto the scene
You can get the same result in normal mode too if you refresh the prompt
refresh as if rewrite?\
No just regenerate lol
...?
You know every time you use the same prompt the results are different
I mean just use the same prompt man how hard is it to understand?
You get 4 different results and all of them look different?
Why are you using just one example?
U need to atleast post 1000 of them
k lol
Then we can get the answer
since every generation is not the same
Or maybe the devs can just tell us if fast mode sacrifices quality?
You know in midjourney the fast mode doesn’t sacrifice quality and the devs werent hiding the fact?
If you mixed them up nobody would be able to tell which is fast mode which is normal
Yeah
Since every video is different
thats true 👌
Can the devs just tell us please so we won’t have to keep debating
ye
yes, there is logic in this, but given the cost and quality of image to video generation, it is unlikely that they would displace them. The power of veo is solely in text-to-video. Well, the powerful whisp is a very cool tool.
Weird! I see fast mode most of the time gives me more dynamics.
Fast
woah was that your own generation?
ok ngl , the normal mode , especially #1 and #2 is exactly what i want omg
from what i noticed your fast mode seems to lean more on cgi/game realistic style
You’re correct, I thought you were referring to the dynamics.
oh no worries, the dynamic matters too
its just weird how both mode adhere prompt differently
ok nvm i guess thats what u get when using fast mode , it reduce the quality a bit , hence the game./cgi render
Is it just me or has the quality in veo 2 become worse?
Have you been using fast mode?
no, I'm not using a normal model, maybe it just seems that way to me
Probably your prompts then
try making them more detailed
Some interesting results from experiments with otter pups. Are they being followed by ghost chickens? 😂
Teleportation is possible:
Strange blinking, with white eyelids:
In the text-to-image-to-video version, they don't blink at all.
This is the best, after I prompted for brown eyes, but the eye movement still seems a bit strange.
Getting better...prompted for brown eyelids.
Lmao! actually fire 🔥 
the quality hasn't gotten worse, that's just how it seemed to me
