#flow-videos・📹
1 messages · Page 24 of 1
we got u btw
I'd like to apologize to the staff for my complaints, I was definitely been proven wrong and that'll teach me to be so overconfident
Yeah, obviously when it’s opening more seat it’s gonna be expected to slow down , I like how the earlier access people “blaming” on us lmao
a lesson on not being so sure I know things
im still tryna get good at prompts tbh
Oops I meant to reply this
yeah
Like tone down the hating, let us use the tool in peace ✌️
ok so, one video generated, the others didnt
who is hating
Well at least you acknowledged 😙 welcome to veo party 🎊
but it may be a censorship thing, my very first prompt was obviously some terrifying cronenbergian monster
🫡 👽 😆
so lucky I got access right on my free day omg
this is not true, its censored at the prompt level then gens are put through a second screening at the end and some videos may not be displayed if they are too violent or whatever
😂 😂
Nice of you to admit your mistakes
i have a midterm in a few hrs and im over here messing around on fx instead of studying
yeah I was very wrong
Bro was here complaining 24h/7
I'm glad I was wrong, lol
lmao
Well lesson to be learned!
indeed
next time I'll learn not to spew judgement fast
I'm very surprised by this, wow
i was too tbh
And now you can see that it IS the best model indeed! hehe
yeah nothing else is even close
y'all know any tool to add audio? I use https://videotosfx.elevenlabs.io/
Someone said also something like MMAUDIO or something in comfyui I dont exactly remember the name
for those than can run it locally
we prob are
there's a hugginface demo for mmaudio that you can try online
Old version of videofx
Mmaudio is I think better
Yeah I second this
letìs see
I wish they make tv mode to play our generated video instead of the placeholder vid
anyone have any good prompts
If you’re stuck, use ChatGPT, or Gemini-
crazy that were getting local models better than those platforms!!
awesome
How exciting
Do you want to try a prompt? Ill run it for you
send as many prompts as u can at this point lol
Lol, if you can, try a meteor crashing into earth. (Just for fun, I wanna see how good it might be at physics)
oook
On it!
More videos the better 😂
It’s more of battle of who prompted it better to get good results
oh, now I can see the progress percentage, I couldn't before
psh i know i didnt
i put "meteor crashing into earth. photorealistic dramatic 4k"
maybe "movie scene intense" ^
Lmao me A meteor crashing into earth, cinematic. photorealistic
"accurate"
then let's add the one-winged angel music theme
i alr know lun sei good at generating prompts
whoops left the oven on
Bro if he put the same effort as the complaints he's level 1000 prompter hahahahha
kelopr what percent u on btw
69
lol
60 for me
can you guys time your gens and report how long it takes
I get a little stuck when generating, can't tell if firefox, internet, or queue
2 minutes average
Trya different browser
from what i’ve heard
yeah I think I may need to
Ive got one generated, waiting for the rest
what browser do u use guys
edge on PC
I'll do it too
If anyone having problem or trouble prompting , google has their own veo official guideline , https://cloud.google.com/vertex-ai/generative-ai/docs/video/video-gen-prompt-guide
havent tried mobile
Bro my meteors look like rockets lowkey but it came out good, waiting for the 2 left
thats what it was before i'm curious if its same now still, or if there are times of day where fewer ppl are using it etc and it speeds up
lmfaoooo
the last two are taking too long here's one
Elongated meteors lol
a better prompt would have def improved results
You got the 4 generated? Im still at 2/4
nah all 4 generated at the same time
3/4
if u got anymore prompts just send them here
Oh okay one got stuck
yeah some dont always gen for me
this doc talks about image to video and negative prompts, huh
Haha I will, thanks! It's 2am here so I gotta head to bed now lol. I'll send some when I wake up tho
cheers
I dunno why I have this fixation on soldiers dancing on alien planets
Im trying one prompt for you hehe
good idea tbh
lol I'm so sorry
Haha, it's awesome! 🔥
Yo mario got them moves!
now I'mma try something difficult: a dude shapeshifting into a beast
Wow!
that poor fish
May I know the prompt that gets you this pov GoPro style result?
From a first-person perspective, the shaky camera captures the chaotic energy of a bustling fish market. The air is thick with the smell of saltwater, fish, and ice. The camera jostles as the person holding it navigates the crowded aisles, the sounds of vendors hawking their wares, the clatter of ice against metal, and the squelching of fish underfoot blending into a cacophony. Close-ups of glistening fish, their scales shimmering in the dim light, are interspersed with glimpses of weathered hands expertly filleting and weighing the catch. The camera momentarily focuses on a bucket overflowing with prawns, their antennae twitching, before panning to a vendor gutting a large fish, the knife flashing silver. The slippery, wet surfaces underfoot reflect the dim light, creating a sense of precariousness. The scene ends with the camera focusing on a dropped fish, its silvery body momentarily suspended in mid-air before hitting the wet floor with a splat, scattering droplets of water and scales.
Comparison with Vidu 1.5:
Vidu is giving my fever dream
Thank you!
nice
mmmh shapeshifting is hard
any tips to make the video look like a- no wait maybe I got it
what kind of shapeshifting?
something like this?
Vidu 1.5 is here! Create your own movies with ease using the power of Multi-Entity Consistency.
Seamlessly blend people, objects, and environments—the only limit is your imagination.
#ViduAI #ViduChallenge #MultiEntityConsistency
I was trying a man turning into some big beast
I may or may not have some weird fixation 🤔
prompt? lol
I’m facing the creative block/prompt block, time to rest 🛌
same. cheers
a question: does this guy look like it's always
the same person?
I genuinely can't tell, I have some prosopagnosia
yes
niiice
uhu I'm struggling to make two characters snuggle. anatomy breaks
lemme add "correct anatomy and poses"
yes, it worked!
3 out of 4 outputs got banned tho u.u
what was the prompt if you dont mind?
I reccommend a less strict filter
From a third-person perspective, the camera focuses on a grassy meadow bathed in golden afternoon sunlight. In the distance, a small, fluffy bunny hops out of the soft, wild grass, its ears twitching and its tiny, twitching nose sniffing the air. The camera zooms in to reveal its soft, velvety fur—light brown with a white patch on its belly—and its impossibly big, round eyes. The bunny pauses, looking up curiously at the man standing nearby. The camera captures the sheer cuteness of the little creature as it hops closer, its tiny paws making barely a sound on the soft grass. The man bends down slowly, holding out his hand gently. The bunny, full of curiosity, approaches with a soft, tentative hop, then nudges his hand with its little nose. The camera zooms in on the bunny’s delicate whiskers and soft fur as it sniffs the man’s hand, before nuzzling it for a moment in pure contentment. The camera catches the man’s face, his expression filled with awe and joy as he carefully strokes the bunny’s soft back with one finger. The bunny closes its eyes for a second, enjoying the affection, then hops around excitedly, its little tail bouncing with each leap. The man laughs softly, the camera capturing the warmth of the moment as the bunny does a happy, bouncy circle around him, its ears flopping with every hop. The camera follows the bunny’s playful movements, the tiny creature hopping in and out of the tall grass, its little body moving like a puffball of joy. The man slowly sits down on the grass, and the bunny hops up to his lap, snuggling into the warmth of his legs. The camera zooms in on the bunny’s tiny paws tucked underneath its fluffy body as it sits comfortably, eyes half-closed in bliss. The scene ends with the man gently petting the bunny as it peacefully rests in his lap, the soft rustling of grass and the warm sunlight enveloping the moment in a tranquil, heartwarming stillness.
cute, I should try some birds
Wow! Thats a big prompt thanks
lol
let's see if this tool knows animals
Man films with shaky first-person camera, walking down a New York street. He focuses on someone happily eating spaghetti candy, their face lighting up with excitement. The camera wobbles as they take a big bite, the candy strands stretching. People pass by, some glancing at the odd scene.
ah!
just use shaky first person camera
nice I'mma try
noice
This video has been added to the audio with artificial intelligence
I wish we could make it longer
guys, I need help with a prompt
the prompt:
Man films with shaky first-person camera, walking inside a dark, dimly lit abandoned house. He opens a door he shows that in the next room there is a female beastly mutant, with shark-like mouth, drool, glowing eyes, sharp claws, veins, crawling on all four towards him. first person view. pov. single unedited take.
the result:
mmmmh
that's one darn good monster but how do I add the dude filming
now im hungry
lol
maybe try
POV, GoPro-style footage. A man cautiously walks through a dark, decaying abandoned house. The shaky handheld camera captures dim, flickering light casting eerie shadows on cracked walls. Footsteps creak on the dusty wooden floor. He slowly pushes open a door. The camera lingers on the rusted hinges before revealing the next room—a grotesque female mutant with glowing eyes, shark-like teeth dripping with drool. Her pale, veiny skin stretches over sharp claws as she crawls forward on all fours, scraping against the floor. The camera shakes as he stumbles back. The scene is a single unedited take, raw and immersive, horror film style.
let's try!
aw hell nah 😭 this one's a different prompt that failed, i'll try the one i just recommended though
she looks cool tho
whats the prompt?
Holy
Shaky first-person camera footage as a man cautiously walks through a dark, decaying abandoned house. Dim light flickers, casting eerie shadows on cracked walls. His breathing is heavy, footsteps creaking on the dusty wooden floor. He hesitates, then slowly pushes open a door. The camera lingers on the rusted hinges before revealing— in the next room— a grotesque female mutant. Her shark-like mouth drips with thick drool, glowing eyes piercing through the darkness. Veins bulge beneath her pale, mottled skin as she crawls forward on all fours, razor-sharp claws scraping against the floorboards. The camera shakes violently as the man stumbles back, his terrified breaths quickening. A single unedited take in raw handheld POV, dim ambient lighting, horror film style.
I'mma lip sync this one
Would eat
@zen nope
I guess it can't associate surreal monsters with a real life recording
kinda like imagen
@manic hedge HELP
this wworked
POV, GoPro-style footage. The shaky handheld camera moves cautiously through a dark, decaying abandoned house. Dim, flickering light casts eerie shadows on cracked walls. The camera slowly approaches a closed door, lingering on its rusted hinges. A hand suddenly enters the frame pushing the door open. In the next room, a grotesque female mutant is revealed glowing eyes piercing through the darkness, shark-like teeth dripping with thick drool. Her pale, veiny skin stretches over sharp claws as she crawls forward on all fours, scraping against the floor. The camera shakes violently, jerking back as the mutant advances. A single unedited take, raw and immersive, horror film style.
you're welcome !! :)
You can use this for Google Veo 2, Runway, Sora, Luma, Kling, MiniMax or YouTube Shorts.
i had to specify like
"okay could you just not describe the man at all??? like it should be a pov gropro footage so it should be a camera following at first and then a hand appears pushing the door and the creature appears."
I just used the general ChatGPT
But I think this is gonna do great
yep! it's been doing me pretty well
hmm i accidentally didnt include the P for POV and it's giving me different results, but i think it still works pretty well
cya!
Oh my gosh this is perfect! It even includes the music prompt too :DDDD
yay :)
Just need to wait for it to generate
I understand why the generation takes so long. It's super high-quality
Anyway ima go
What could this fella possibly be contemplating
Your Turtle scenes are cute! Did you mind sharing the prompt?
From a third-person perspective, the camera captures a man standing by a serene, grassy riverbank on a bright, sunny day. The peaceful atmosphere is interrupted by the soft, unmistakable sound of tiny paws pattering on the ground. Out of the water emerges a small, fluffy baby turtle, its little legs scrambling in excitement as it moves toward the man. The camera zooms in to reveal its soft, textured shell, bright eyes, and curious expression. The man crouches down, a gentle smile spreading across his face, and extends a hand toward the turtle. The camera focuses on the tender moment as he carefully picks up the tiny creature, its little head poking out of its shell in a display of innocence and curiosity. The camera shifts to capture the man’s face, his eyes full of wonder and warmth as he gently cradles the turtle. The turtle, feeling safe, begins to crawl up his hand, and the man lightly chuckles, amused by its tiny, determined movements. The camera pans out to show the surrounding environment—a tranquil river flowing gently in the background, soft grass underfoot, and wildflowers dotting the landscape. The man carefully places the turtle on the soft ground, watching as it moves with slow, deliberate steps, exploring its surroundings with a sense of newness and wonder. The camera captures one final shot of the turtle waddling back toward the water’s edge, its adorable form highlighted by the glistening sunlight, before the man watches it disappear into the river with a satisfied smile. The scene ends with the quiet calm of nature settling back in.
Thank you!
I love how it changes the shot
I still think these ai video gens are really stiff when it comes to anime
Anyways, does it know tf2? Can anyone try it?
Depends on the video model
Vidu 1.5 is the best model for anime stuff
Yes
Its taking 6-7 minutes to generate right now for me. Do you guys have the same time?
The generation time is taking longer time than yesterday yeah
It is longer this morning
yeah lol, i used a simple prompt to test it, it can probably look better than this lol
Fortressnite
😌
Yeah
At least the text wasnt like "GakD MoNriNG" AI usually does that
I hate when the percentage thing just dissapears though
same
Lol trust me I've been doing Good morning slogans for quite awhile now. I know
Opposite day
You can but avoid posting that here
Ahh okay
Jane on my Mind https://youtu.be/Ho0-RXdk9mI?si=kKXJWNt61fRqPNpD
#aifilmmaker #aishorts #aicommunity #cinematic #gamiorbegoso #veo2
that looks like the same guy, how did you do it????
guys why do all my video look ultramegahd, it makes them less real-looking imo
Always use the same prompt to describe the guy
how detailed must it be? can you give an example
I'd really appreciate, making a long video would be so cool
yesss I love alien planets
thanks!, " Mike, 18, slim, bleached-blond hair"
that's it? maybe it's the name giving, lemme try
Does anyone know how to fix the scaling issue on mobile? It’s stoping me from generating on mobile
Still looks decent
try this "Dynamic tracking shot: Mike, 18, slim, bleached-blond hair, speeds down a dusty road on his yellow motocross bike,at night, illuminated only by the street lamps on the highway His Sony Walkman dangles from his jacket pocket, cassette spinning inside Low-angle shot, at one point Mke closes his eyes and opens his arms..The scene has a raw, 90s indie film aesthetic with a slight film grain texture.
Looks nice
Did you be very detailed with it?
Lemme try to depicit it
An elephant + a deer + a um... dinosaur
Ooh! and a Gorilla
Hmm looks like firing nerouns in the brain but the nerouns are clouds
SK you also got access?
Lovely transition by the way half-way in the video
Oh
What did you originally expect?
Okay
im enjoying veo 2 sooo much
Ahhh red lightning
Looks like something SUPER bad is preparing to happen
Ahh well anyway Ima have more fun with Veo 2 😄
Byeee!!!
👋
Sorry i don't get it. So they can give you access and revoke it after a few months ?
Sorry, I thought you meant you were given access that would last for a few months. But I get it now—do you mean that you got access a few months ago?
English isn't my first language.
man why it doesnt work for me mmmmh
Not this morning, speeds were don much slower, I assume it was cause soo many people were using it
Just needs to learn how to count
don't judge me, this was the plot of a compelling scifi short story
she becomes infested by the aliens (cliché on the surface, I know)
but what matters is that the alien is a lot like I envisioned!
now to find how to make a dude turn into a beast
Why are so many people having bugs in Veo 2 I'm chilling 🏖️
I'm not having issues except for the censorship, works fine
hey did I accidentally manage to make the same woman??
I should be like Chris Cunningham and manage to tell a whole story in 8 seconds
that'd solve my consistency problem
Aww what a cwute wittle bwaby :c
fluff
uhoh they seem plastic dolls. any tips to fix?
there's some specific prompts that make the ai want to use a certain style
try using photorealistic in your prompt
Yesss! The first time I tried it, I got like 3D cartoon animation generations
Hoo you got your access Lun-Sei?
Congrats
yessss
dont gotta flex that sword 😅
Dang this is work of art
you told a whole story in 8 seconds. I'm amazed.
Indeed I have
why am I so verbose, I -must- find a way to tell a story in 8 secs
my videos are always like "and this is like a tiny moment of the story I have in mind"
You have to speak to it as if that person is human
such as?
Reason I have good prompts and images is because I'm a poet and artist altogether. So language isn't an issue for me. Far as feeling out a particular scene to convey
my main issue is being able to do a full-story in only 8 seconds. some people manage in 5, but I'm just long-winded by nature
You just have to keep building on experience...then you'll cook!
I use Fal, sorry
on the plus side I can make random dances
That i noticed
I can make random people have supernatural powers 😄
A skill trait uniquely based on you
Yep
I have some many gens I made. I'm just curious to see what you guys make next
my issue with it is that I need character consistency
Oh wow 👏
Doctor Manhattan sister
It only generated one of the water generation
Love how no one is reacting to it
They are just looking at their phone lol
Wow surreal visual effects 🤯
nice veodeos y'all
Im gonna generate a video of a dog after this fantastical video finishes generating
Anthromorphical dogs!!!
The 2 back dogs are very excited about what they are gonna eat
Oh my gosh it turned out so GOOOOD
Frfr
Comedic timing experiments continue.
Oh, now that you're doing this, can you try the typical Superman scene where he runs and opens his shirt with the superman symbol etc?
Oh oh??
you mean actually superman?
Actually superman
I'll try now
@buoyant merlin didn't work, I got one man as clark kent and another as superman.... lemme retry
darn thing!
@buoyant merlin
must be a prompt issue, not the first time it happens
Who tf is this 😭
apparently he's superman @buoyant merlin
Which prompt are you using? I'll try to see if nim.video improves it
clark kent, wearing a simple business suit, runs with a focused expression. he turns the corner and opens his shirt to reveal- WAIT
Looks awesome, mind if you share the prompt?
maybe it's "shirt" the problem
"opens the jacket to reveal under it there's the superman symbol"
oh it was merely "a pigeon dancing in a tutu, cute, cartoon"
This one i generated with Nim video's Veo
Why are they praying to that quasi-triphobic beast?
Hmm lets see. A creature mixed with a Collosal titan + dragon + leopard + whatever animal head that are on the 5 heads.
@buoyant merlin this is getting irritating lol
It's a verse from Revelation
Ahhh I see
Bro I just sent like 2 messages into a GPT and now they say I hit the limit
ChatGPT is money hungry you gotta pay to attach a dang image to it
It definitely struggles with men experiencing profound grief. Also if I denote the person as wealthy instead of poor, they become more plastic.
Grok is far better than gpt, anyway
Psst - #todays-special・🥣 is up.
you beat me to it~ we're celebrating national strawberry day, if you need some veo inspo
🍓
#todays-special・🥣 message
Whats the prompt
once more, I wanted real life and I got videogame/cgi.... what do I do wrong?
I generated it with ChatGPT lemme try to find the prompt
( grok > gpt )
I mean like I didnt generate it with ChatGPT I generated it with Veo 2
but i generated the prompt with chatgpt
Do you have a Grok prompt generator prompt for Veo2?
@pseudo cedar 'Amidst a serene, dew-kissed meadow, a flower god stands in the heart of nature's embrace, their form radiating an aura of vibrant life. The camera glides in smoothly, capturing the delicate sway of grass and the soft glow emanating from the god’s presence. With a graceful motion, they raise their arms, and streams of vivid, chromatic petals rise from the earth, swirling around them in an enchanting spiral.
Petals of pink, white, red, and purple dance through the air, creating a cyclone of color that moves with fluidity and purpose. The camera sweeps around, offering dynamic perspectives as the floral whirlwind grows tighter and faster. Each petal glows with a gentle luminescence, leaving trails of light that paint the atmosphere with a dreamlike quality.
With a sudden, powerful motion, the flower god thrusts their arms forward. The floral vortex compresses and then erupts forward at super sonic speeds, a cascading wave of petals and blossoms that tears through the air with a soft, whispering rush. The blast’s impact sends a rippling shockwave through the landscape, and wherever the petals touch, vibrant flowers spring to life in a breathtaking timelapse effect.
The meadow transforms into a tapestry of blooming beauty, petals unfurling in rhythmic waves as nature itself awakens to the god’s touch. The camera pulls back, revealing the sprawling sea of color stretching to the horizon. The flower god stands at the center of this blossoming paradise, their silhouette framed by the endless bloom. Soft winds lift petals into the sky, creating a mesmerizing display as the scene fades, leaving the world awash with life and color.'
haven't tried because I usually write the prompts myself, but you can just ask it I guess
Thanks
You're welcome
slightly more real-looking now
even more realistic
but that mutant's muscles need to be BIGGER, absolutely disproportioned
epic
I know Poe, etc has Veo 2 but do any third party places have image to video yet?
Freepik has it
I sure love posting videos to the video gallery channel.
Really good videos!
You could probably make stock animal footage
Hair movement looks awesome on subjects!
Holding focus without jumpcuts
Blue eyes white dragon
i owe all the video credits to Friends on Here prior to me getting access (only uploaded to my personal-youtube-unlisted)
omg. please share more details on the ants and inspo behind it!!
that was my opening prompt lol
and then i iterated
lol yes, and many many more 
Would love to know the prompt for this one
Man films with shaky first-person camera, walking along a rocky shoreline. The camera tilts down to a sea cucumber lying half-buried in wet sand, its gelatinous body glistening in the sun. He crouches, fingers hesitating before gently poking its squishy surface. The creature wiggles slightly, oozing out a strange, translucent substance. The camera shakes as the man chuckles nervously, then slowly pets the slimy, bumpy texture. The sea cucumber pulses faintly, as if acknowledging the gesture. The man holds his hand up to the camera, fingers coated in goo, whispering in disbelief before rinsing it off in the shallow tide.
@polar hollow crushes it
Floppa on a floaty
death stranding?
50 points to whomever can make a sock puppet that isn't constantly flapping its mouth.
Indeed
Sadly "death stranding" as word gets blocked though
What did you go with - kojima?
What finally got there?
I sadly dont have the original prompt anymore
But I can try re creating it with another similar prompt
I also use AI Studio; it’s much better than chatgpt and the others.
when i use Veo2 specifically in my prompt, ive only used AI Studio, but i tell ppl with Sora to try both oAI and AI Studio 
The scene opens with a low-angle, extreme wide shot of a desolate, rocky landscape, directly inspired by the stark beauty of Iceland, the entire area shrouded in a thick rain. The lighting is muted and diffused, creating a somber, oppressive atmosphere that reinforces the feelings of isolation. The camera then executes a slow, deliberate, smooth tracking shot, following a lone porter, clad in a worn, futuristic delivery suit, their silhouette stark against the grey sky, as they begin to move forward with many packages on their back. The figure is framed using the rule of thirds, emphasizing their isolation and the immense scale of their task. A smooth, slow rack focus shifts from the rain-soaked rocks in the foreground to the porter's face, revealing a weary, but determined expression. The scene uses a shallow depth of field, blurring the distant landscape and highlighting the porter’s struggle. The camera then cuts to a tight close-up of the porter's hand, gripping a worn container, the texture is ultra-realistic, emphasizing the worn-out feel of the equipment. The camera executes a very fast and dynamic whip pan, the scene quickly shifting as the porter suddenly stops, sensing a nearby presence. The sound design includes a combination of the distant rumble of thunder, the soft patter of the rain, combined with a low, ethereal hum that grows increasingly louder, giving a sense of approaching danger.
Thank you!
Warming up on set of their first workout video. 🦾
@drowsy nacelle I really need to learn prompting from you, it’s really good how consistent most of your generated video are
Made another one of these
just use AI
Thats what I used to create that flower exploding thingy
Looks great!
What do you use for audio?
Would you mind sharing @drowsy nacelle
MMaudio
Thanks I appreacited
It turned out even better! thanks @drowsy nacelle for telling me about MMaudio
The other MMaudio SUCKS. It makes me pay (which I don't) before I can use it
yeah it gives me a paywall too
:/
Yeah like Sora
VideoFX however 👼 . near-infinite generations, high-quality, and no cost
Yeup thats a w in my book
Oh and those claim to be free generators that offer free generations but then you see the credit currency UI 💀
Use hugging face one
Yeah thats what I was using to create my other AI generated videos. But when I first used the other MMaudio, that paywall annoyed me
Yeah it caught me off guard too when it suddenly
Shows the payment plan lol
If anyone interested, https://huggingface.co/spaces/hkchengrex/MMAudio it’s to produce audio from
Your generated VideoFx video it’s quiet easy and straightforward to use, u can leave the
Prompt blank or whatever you desire ,
Wait a minute
You can leave the prompt blank?
Yeah
Even in the hugging face one?
But , it’s optional , if you want more accuracy u can prompt a word or two
Yep, I mostly tried leaving it blank to see if it matches with my expectations
I've been typing whole sentences 💀
Oh lmao , that’s fine too , I mean if it gives you accuracy long sentence are fine I guess
But if you look at examples
All of em uses one word or two
Kingdom Hearts Union X 2 looks cool XP
That's not official, and doesn't have anything to do with the actual one and is NOT affiliated—check their github.
Either make a new account if you are unable to use it or just wait 24 hours.
i think Veo2 will be huge for fashion, heres a quick example:
"professional runway fashion show, a woman wearing a giant mushroom hat, velvet cape with embroidered stars, asymmetrical striped trousers, curled-toe boots, full-body shot, dynamic angle"
Isn’t that aurora
lmao true
lmao love this one
Omg they’re either hiding or frozen to death
they sneak
Is it just impossible to describe how great it is! People won't understand; is it too revolutionary, like something we have never seen before in our human history.
It wants to make temple run mobile games so this has a lot of manual prompting for cam and movement.
Thanks Google; going to tell my friends this was me 
First person E.T. when
Have to agree it's revolutionary; I used it all the way back
Generally I agree that it's a better idea if you know what you're doing, though it can be tedious especially with sync compared to just throwing it into a service. Plus, there's many misses with MMAudio. As always, I wait for when Google's official version of this goes public https://deepmind.google/discover/blog/generating-audio-for-video/
There's a lot out there that can actually sync up the mouth movements for you
MMAudio runs great with Pinokio locally
haha I just use the HF space
Gosh that's amazing
Ahh so when did Imagen 3 learn about gravity? 🫨
Use huggingface
I do. I was talking about the other one
TW: worms
oh nahhh 🤣
heh
Toxic Reign EP, Track 02 - 04: GHOSTS IN THE RUST (ORIGINAL VERSION) 02:47
Just... AHHH I love VideoFX
It can do almost everything, it can generate almost everything, almost high quality, almost infinite generations. And its SUPER free
Remove all those almost from my sentences because this is FIRE
i mean the fact AI has come this far is absolutely insane... i remember 2021 when GPT-3 couldn't even answer basic questions
Oh dang I tried to generate this but the "cute girl" in the prompt got flagged
For real
Why use Suno and not Udio 😭
and to think this is still in it's early stages
Thank you 🙂
suno can be considered superior imo... they have been leading the path for AI music generation for a minute now
obviously it's subjective though. haven't tried udio in a minute so not sure what they did to improve
I’ve been using Suno since Version 0, it has gotten a lot better and is very good, and Udio is very good too, but I just think every Ai Video that uses Suno is brainrot because of the quality in the Audio
ur right... tbh i wouldn't consider either for these purposes

u know what i can't wait for though? the next generation of musicfx
Hexagon
i swear musicfx even a year ago was better than its current state
decided to continue this first generation and add music, I really like the result
Gosh I said that I was gonna make one more generation before I go but here I am making another generation 💀
This would make a pretty cool discord profile banner
love this onee
damn veo never fails to impress
so grateful we had the opportunity to gain access, thought it would never come 🥲
pshh if only temple run actually looked that good
lol
Is anyone gonna give access to Veo2. I'm dying to try it. Lol!
The videos here are making me jealous.
i feel like veo 2 is the type of thing where you use it for a few weeks and then forget about it
You can go to freepika to have a try
send a prompt and i'll get it started for you
by the way.... the only reason i got access recently was because i signed up through the form in #flow-announcements・📣
the form is closed now, but make sure you follow it incase a new opportunity comes
I beg to differ sir.
yeah really depends on your use case
Ok sir. I appreciate. I'll send but I really hope to get access cuz I have like 10s of 1000s of prompts. So I really need access so that I can generate it whenever I want to.
Drop prompts here or your DM? They are 5 can you? Or only one prompt?
follow that channel, it was only a couple days ago where they started allowing signups. unfortunately it's not open anymore, just make sure u follow the channel
send them here. as many as you like
I do sir I've been following it closely unfortunately that day I discovered a music Generator somewhere and I was busy generating songs. I didn't open my discord that day.
Ok sir. Lemme craft the prompts and drop them here. Thanks.
curious to see what u come up with
🙂
just a tip i'll send here that i found to be useful
Currently working on a short movie about an imaginary African country.
sounds fun. feel free to ping me with a prompt if you make one, i'll be around
Ok. Working on the prompts.
Why can I generate on veo 2 after reaching 100% it doesn’t show me a thing
Hello sir. I'll drop prompts on #flow-tips-n-tricks・✏️
do it on PC bro. it has issues on iOS
Same thing on my pc
what browser
Chrome
try edge
Alright
just to make sure use the "text to image to video" and generate the shot to see if that even shows up
on edge
lmk if it works
does veo 2 have a word limit?
Yes, I dont know exactly what is the limit though
I think It's similar to the imagefx one
But it’s better with medium clear and concise prompt
It's honestly crazy to have a model like VEO2, with such coherence and prompt following only like 2 "real" years into the whole AI video stuff. I mean since they've been really pushing the ai videos space, such quality in such a short time..
Maybe it’s just me idk
Definitely not for me! lol
I think try to think outside of the box when it comes to prompting what you want !
You’ll never be bored
mhm
If you have real use cases for it, want to create a short film or whatever you'd never want to stop using it 😉 Especially VEO2 because It's so good
I think what makes it fun for me is when I don’t look at others people generated video and comparing with mine ,
It’s the result that’s excites me
Yeah just letting your mind looose
Wan 2.1-veo 2 comparision
In comparison to what? To real life, sure. To other videos models, It's still better than them 😄
remind me of this
true, wait i wonder if veo2 do well with punching bag
kling 1.6 does well
Yes It's just I've been trying to generate battle and fights before with other models that I feel so happy now with veo2 fighting even if Its buggy 😂
this scares me in a funny way u can tel that veo realy wanted to let that first strong strike to hit him directly
wan 2.1 got the potential, from the looks of it
suddenly BOOM veo 3 announced
Lol
the car door being slight open after the man got pinned down 🔥
This ones pretty good too, wan or veo?
thats veo
Yep!
Bro I can't I remember just 6 months ago I thought it would take years for AI to come up with semi-good fighting like that 😆 AI goes so fast
Yeah they been creating some good stuff with VEO2 already
ITS SO GOOD
A sci-fi type shortfilm
They really make the model shine yep
yeah i think veo2 got that slight hit and miss with interacting with punching bag
wait actualy
this looks good yeah i take it back lmao
I dont understand why veo 2 suddenly changes the frame suddenly
Yeah it's similar to stock video where the boxer doesnt have any technique, but at leasts its not bugged out! 😄
Nope the technique is not good
anyone wanna give that a try
But it's like a beginner
or find a loophole to get veo 2 to generate tasmanian tiger
It's because the training data which involves stuff like movies etc is often with cuts. When you watch a movie there's a lot of cuts to different angles etc
I had some success including this #flow-videos・📹 message
You are right, but i mean its a little frustrating you cant change that, even you tell him to not do that
to avoid cuts
oh i need to keep this in mind
Did it work? It didnt for me
Keep in mind it doesnt always work, and the content of the prompt is important too. If you had too many components etc it will try to cut
to fit that 8 second frame time
honestly with veo 2 , its more of you learn what line did model decided to use and ignore
trial and error thingy
If you combine it with a prompt that doesnt describe too many stuff happening at the same time (since It's only 8 seconds) it works moderately well
and also the flow of the prompt like what do you put first in order
This!
but based on google officia veo guideline
oh sorry i meant this
this is how it would prompt in following order but i dont think it matters much
I soooo need to add audio and lip sync here
prompt??
prompt?
one moment
A brutal street fight unfolds between two men under the glow of flickering streetlights. Punches land hard, bodies stagger, and the sound of heavy breathing fills the tense air. The camera moves dynamically, capturing each strike and dodge. After an intense exchange, one man grabs the other with force, lifting him slightly before hurling him backward. The body crashes violently into the side of a parked car, denting the metal and shattering the window. The impact echoes through the empty street as the fight momentarily pauses, both men catching their breath.
dm
@sly glacier We need to control the cuts, any idea how to do that?
I'm testing also but as you know its trial and error ...
Yup.
I get like 70% of videos without cuts personally
from what i notice, veo 2 and sora have similarities in term of doing things we didnt ask
I often add the bit I've shared in the prompts, not sure if Its this or the overall way of prompting that helps it
sora has it worse tho
Oh I see
70% should be enough to find great outputs though, not that much of an issue
safe to assume that .. the model just generate what it thinks best hahaa
what matter is that it follows the subject in our prompt
i do wish it can follow the prompt consistently for camera motion style
god im very much amazed how good imagen model is too ahaha
My experience over time has been prompting "continuous shot" in text to video does mostly result in continuous shots. Text to image to video seemed really prone to cutting away to random unprompted shots. That may or may not be the case over the past month, but I personally wouldn't know, because the experience led me to focus on T2V as I really mostly care about continuous shots I can edit myself, except for the occasional single generation meme use of the tool, which is the only real time I want to control a cut within a prompt.
same
oh yeah veo 2 I2V more prone to getting cut
text to image to video is either ... confuses at reading prompt or im the one being stupid cuz im pretty sure the prompt i wrote was straightforward
the sudden cut to shot of different thing caught me off guard
like the model managed to actually add motion to the image in first 2 seconds then suddenly , cut to different things
Yeah, in an ideal world I would use T2I2V to quickly confirm that I was on track to get characters/aesthetics/environment/blocking right. But it simply doesn't feel worth paying the "random cuts" tax. But I try to sneak back and check it out from time to time to see if the model has improved.
Hmm seems people asking how to keep a continuous scene ...well 8 seconds isn't a lot to go off from. Considering the amount of depth added
its more of just one shot instead of cutting into scene thats not relevant with the user input i supposed
Plus it varies on the prompt suggested
yea its kind of hard to predict, but i guess thats why they gave us four variation in one prompt
Yes best case scenario
at least with videofx you can still try and error, mean while other ai video generation model doesnt do that, it sticks to your credit use, and if it fails , they wont even give u credit refund
which pisses me off with some of the ai generation out there
You could locally run hunyuan using comfy ui
Or wan 2.1
Have you tried wan 2.1? How’s the experience!
this is a great comparison example
Nope, just heard that it was open source
Y s, correct.
decided to drop it now
probably they saw google giving veo api access to most people
@muted quartz made this one pretty cool but it's kinda too fast haha
btw you can't make other Jaegers XD the model is so biased that it can only produce Gipsy
Testing it out for creating some cool b-roll footage; so far it's beating expectations . 
AI is just truly incredible
I love how well these are done. I wanted a CGI or realism, but it did anime which is still cool.
what was the prompt if u dont mind
I already erased for a new prompt for a video
Something like high quality cinematic a short white hair woman with red eyes. Standing on top of a massive mountain looking below at purple leaf forest and field with tan grass. Red star making the sky red.
When are we going to get a bird sim game
hey,
oysters
how do you do transformations in veo2? I don't think veo2 is capable of doing that
A self driving Tesla
reflection test 1
What I imagine is how it should be spelled the language he is speaking
Escie adahot sebere oquilas. Hieye frieby
2
has anyone managed to do a shapeshifting sequence?
Can you write in more detail what you mean?
a man turning into a beast, or like a pop singer turning into a d33mon
I'll try this and write to you later, my battery is low
ok
be careful, it aint easy
The fox king!
Father and daughter sparring
I sort of tried this with a woman diving under a wave and appearing on the other side of the wave as an otter and snail. But, instead of diving, she jumped into the wave. #flow-videos・📹 message
shapeshifting doesnt work apparently
dang
"Take care of Mr. Fluffers if you know what I mean."
Lol see you later alligator
In a while crocodile
Lol
because why not lol
prompt???
Master Chief in Mortal Kombat fighting Pennywise
Is there an easy way to try Veo2 from Canada?
does it know league of legends
Likely it's a famous game
As long as it's a huge massive well known one it should
2D videos also nice ))
Nice walk cycle
little bit flickering, but i thing few rerolls will fix
Possibly
An attempt at the ring
how? ))
@old frigate what you mean?
jellyfish )) Because all my jellyfish's same ))
creppy )
Ooh you have to describe it the way you want it
Sorry I can't put the prompt because it has the forbidden color
Bioluminescent jellyfish with translucent hues glowing etc
A dark forest with a blue tint and a glitchy video tape effect filter. In the middle of the frame is a little stone circular well with no roof. Camera gets closer to the well slowly to show a long dark color hair girl with a dirty white dress crawls up out from it. Her long dark color hair covers her face and falls down in front. The Ring
Changed to dark color
Mafia cat
I tried every variations of jellyfish's, same result )) haha
@manic hedge Was this as you wished?
Moonlit people, a race of zombies basically.
@manic hedge First make a normal man and then tell how the man gradually becomes what you want.
Use my prompt
Bioluminescent jellyfish with translucent hues glowing etc
tried, same )) But anyway, thank you ))
more or less, yes!
what was the prompt?
Unfortunately I deleted Promt.
First I drew the picture, then I wrote that the man's horns and face turn red and it becomes Devil
If you used VideoFX, this has a record of your generations
What I wish could be fixed is truly understand a style. I wanted anime, but it did something like unreal engine 5. I love this style and my favorite.
I was wondering how to access the VideoFX library - is it possible?
There is no video fx in my library
But it also actually followed and did anime
When will Veo 2 be available to everyone? And what is its pricing like?
Oh, you're right. Strange
The anime when it works and is good is awesome
what app?
How do we get to the library?
There is no library, he just thought so
What is the pricing for Veo 2? Can't they offer a subscription package model?
Where is the secret door? 😂
There's a dropdown on the right
That should show VideoFX, but they may not have it enabled yet
Videofx is not in the library right now, it will come in the future
It's not enabled in my library dropdown.
Why is Poe not recommended?
Then they don't have it. It should be coming soon™️
Write your request here so people can do it for you
So expensive, I really can't afford to play with it.
If it could be like OpenAI, $200 for unlimited use, I would pay without hesitation. Unfortunately, judging from the current pricing, it doesn’t seem likely.
do you realize openai's sora is trash
Sure
well then
I have never used Sora.
However, judging from Poe, it seems to be much cheaper than the pricing on Vertex.
Seeing this pricing, I’m just speechless and devastated.
I don’t know when the day will come when ordinary people can afford it.
For example, when there are some shots I don’t want, and the prompt can’t precisely control the visuals, I might need to generate multiple times. In that case, I think this pricing might not be lower than that of a professional team.
Such expensive pricing feels like the price of a Bugatti.
Although Sora is terrible, seeing this outrageous pricing, I almost feel like I can accept how bad Sora is.
I’ve been waiting for Veo 2 for over three months and haven’t used it once. By the time I found out its price, it was always something I couldn’t afford.
Shape shifter ?
we're diving under the sea for today's special!! 🪸
#todays-special・🥣 message
hahaha that was so unexpected
I just tried using Sora, and it’s honestly trash. Is OpenAI not planning to update Sora anymore?
that's very good! but she's not shapeshifting
told ya
gonna keep working at it... getting closer
So what’s out there on the market that’s better?
I waited three months, and the result I got was 50 cents a second.
This is absolutely fantastic news.
Tried to do like a dyson ring I guess around the star.
If it were you, would you pay 50 cents per second?
Can you select time and resolution on Poe?
Anyone knows where the downloaded video goes on iphone
in your files-- download i think
1080p! we dont have that one in the google labs version
i want to see how good it looks
if anyone generated one feel free to share
woww
Its not there
A wide shot of a grizzled desert wanderer with weathered bronze skin and sharp amber eyes, his face half-hidden beneath a tattered hood stitched with faded runes. He clutches a gnarled staff
of course, just want to see how it looks
Wait there's even negative!!
Let's just put "low quality, artifacts, illustration, 3d" and 7 seconds. thanks a lot btw!
😂 🔥
I see, np
thanks
-
Lens Focal Length
- Focal Length: 24mm (full-frame camera)
- Analysis: The image showcases a wide view, including the entire village, fjord, and surrounding hills, with no noticeable edge distortion. This suggests the use of a wide-angle lens, but not an ultra-wide one (e.g., below 16mm). A 24mm focal length provides sufficient depth of field, capturing the expansive scene while maintaining a natural feel, making it ideal for simulating the perspective in this image.
-
Shooting Style
- Style: Landscape photography style, aerial perspective
- Description: The image is shot from above, offering a majestic view of the fjord, green hills, and the village’s overall layout, as if captured by a drone or from a mountaintop. The lens perspective is balanced, with the fjord serving as a central horizontal line guiding the viewer’s eye, flanked by green hills and the village along the water’s edge. The cloudy sky adds drama and tranquility. The depth of field is deep, with both the foreground (village and grass) and background (distant hills and sea) clearly visible, indicating the use of a small aperture (e.g., f/8 to f/16).
-
Filter
- Filter Type: Polarizing Filter
- Effect: Enhances the contrast of the sky and water, reduces glare on the sea surface, and makes the green hills and blue waters more saturated. The sky appears in soft gray tones with rich cloud details, the water shows a deep blue, and the grass looks vibrant and lively.
- Additional Adjustments: A slight cool-tone filter may have been added in post-production to emphasize the overcast atmosphere, further enhancing the drama and mystery of the scene.
Bro, can you help me out?
When you go to download it should pop up at the bottom of the screen where you can choose to save it and then it will ask how you want to open it, i usually click open in and send it to myself on Facebook to save it to gallery from there
Also I’ve noticed sometimes it refuses to actually download on iPhone, my fix was chrome had to update lol
I can only describe it with a “holy st.”
I can only describe it with a “holy sht
thanks,At this quality, I’d be willing to pay 50 cents per second.
every now and then you return begging someone or some company to take your money. I can't tell if you're a troll or genuinely have issues, but stop pls
So you even know it’s me begging others to take my money? Someone who didn’t know might think I’m the one trying to take your money!
I’m just discussing the pricing—if you find it annoying, you can hit the “mute this person” button for me.
you dmed me privately asking me to sell you my account.
thanks! looks good 🔥
and every now and then you come here yelling about how much you want to spend a lot of money. It's getting on my nerves
we have negative prompting??
You can choose to sell, or you can choose not to sell. But how long ago was that?
Can we stay on topic guys?
Guys let's just calm down lol if anyone got a prompt I'll run it
I assume kg was talking about another video gen with negative prompting cause videofx doesn’t have it right?
hey y'all! this is a place for sharing videofx generations & collaborating. please keep discussions either in #off-topic・🐢 or in personal DMs. please also keep in mind the #rules-and-guidelines・📋 before engaging. thank you!
Pin that please.
📌
SK has access to negative prompting through Poe! Apparently we can do that there that's what we were talking about
Could just disable text comments and enable only uploading videos
Oh?
first of all, I’m not showing off wealth; second, I’m expressing that 50 cents per second is really expensive. I want to spend money on this, and it seems to have made you uncomfortable. Are you from Nigeria? Does even this amount of money make you uncomfortable?
running
thanks, always nice to feel appreciated
the dream is that we can share collabs and add comments in a respectful and productive way, but this will be what happens if things change. thanks y'all
trying to do some parkour
Pretty good fight video