Veo 3.1 AI Video Generator
Google’s video model, for realistic scenes where people speak and the world makes noise, in one pass.
Write your idea first. You sign in when you press Generate.
Short answer: Veo 3.1 is Google’s AI video model, introduced on October 15, 2025. It makes realistic video together with dialogue, sound effects and ambience, follows prompts closely and turns a photo into a clip that keeps it. Genvima puts Veo 3.1 in the prompt box above, so there is no Flow seat, Gemini plan or Vertex project to set up: type a scene, add a photo or two frames, and receive a 8 second 720p or 1080p clip with sound, watermark-free, a few minutes later. Veo 3.1 starts at 200 credits a clip and Veo 3.1 Fast at 28. It is paid with credits; to try an idea first, the free Basic engine in the same menu gives you one free video and three free images a day.
How to use Veo 3.1 online
You do not need Flow or the Gemini app. Write here, sign in with Google when you generate, and download the result.
- 1
Choose the opening
Start from a written scene or a photo that becomes frame one (Text/Image to Video), or lock both ends of the shot with two stills (Frames to Video). For Reference to Video, Google’s "Ingredients to Video", switch to Veo 3.1 Fast.
- 2
Write the scene, the line and the sound
Describe the shot and the camera, then put speech in quotation marks: She says "We open at nine." End with the sounds you expect to hear.
- 3
Choose Veo 3.1 or Fast and the resolution
Every clip is 8 seconds at 720p or 1080p. Veo 3.1 is 200 credits at 1080p; Veo 3.1 Fast is 28, so draft on Fast.
- 4
Get the clip
Veo renders in the background. The MP4, sound included, is saved to Assets, so a closed tab loses nothing.
Veo 3.1 key features
Official reference clips from Google DeepMind alongside our own Veo renders. Suggested prompts are separate from the clips. Vendor showcases may use settings and inputs beyond those offered here.
A raccoon talks in a cafe, animated from a reference image. Try a related prompt
Two old friends sit on a park bench in autumn. He says "You still owe me ten dollars." She laughs and says "Twenty years and you remember that?" Leaves rustling, distant children playing.
Dialogue, effects and ambience in one pass
Google calls the audio in Veo 3.1 richer than before: speech that matches the lips, footsteps that land on the beat of the walk, a room that sounds like a room. You write it into the prompt and it comes back mixed with the picture.
A camera moves through a sunset kitchen toward three drink cans, combining product and setting references. Try a related prompt
Handheld shot in a small bookshop on a rainy afternoon, an old man climbs a wooden ladder and pulls out a red book, dust in the window light. Rain on glass, a creaking ladder.
Realism and close prompt adherence
Veo 3.1 follows a detailed prompt closely and renders realistic textures and light, which is why it carries the "Best realism" tag here. If the prompt says the cup is chipped and the light comes from the left, that is what you get.
A singer performs against a colorful floral backdrop, guided by character and style references. Try a related prompt
Animate this portrait: she looks up from the camera, smiles, and a strand of hair blows across her face. Soft street noise, a passing bicycle bell.
Photo to video that keeps the photo
Google says image to video improved in Veo 3.1: the first frame stays faithful to your picture and the motion follows your prompt. Use a portrait, a product shot or a painting as the start.
Our existing ice-cave shot, made with Veo 3.1 from start and end frames. Try a related prompt
Move from the first frame, a dark ice cave entrance, to the last frame, a bright frozen lake under the aurora, flying forward through the cave. Wind howling, ice cracking.
Frames to Video: a start and an end
Give the first and the last image and Veo 3.1 makes the shot between them, with sound. It is how we made the loop on this page: six shots, each with a fixed start and end.
A miniature castle rises from latte foam, combining coffee and architecture references. Try a related prompt
Use @image1 as the chef and @image2 as the kitchen. He tosses vegetables in a flaming wok and says "Thirty seconds, no more." Sizzling, a roaring burner.
Ingredients: your own characters and objects
Google’s Ingredients to Video builds a clip from several reference images for characters, objects and style. On Genvima this is Reference to Video on Veo 3.1 Fast, up to 9 images; name them in the prompt as @image1, @image2 and so on.
A Veo 3.1 Fast ice-cave shot from our homepage film, shown here as a short excerpt. Try a related prompt
A golden retriever runs along a beach at sunset and jumps into the waves, the camera tracks beside it. Waves crashing, happy barking.
Veo 3.1 Fast for drafts and volume
Veo 3.1 Fast takes the same prompts at 28 credits against 200. Use it to find the shot, then render the version people will see on Veo 3.1.
Veo 3.1 specs and credits on Genvima
| Model | Best for | Resolution | Clip length | Sound | Modes | From |
|---|---|---|---|---|---|---|
| Veo 3.1 | Best realism | 720p, 1080p | 8 s | Yes | Text/Image, Frames | 200 credits |
| Veo 3.1 Fast | Faster | 720p, 1080p | 8 s | Yes | Text/Image, Reference, Frames | 28 credits |
| Seedance 2.5 | Top quality | 480p, 720p, 1080p | 5, 10, 15 s | Yes | Text/Image, Reference, Frames | 97 credits |
| Kling 4.0 | Cinematic | 720p, 1080p | 5, 10, 15 s | Yes | Text/Image, Reference, Frames | 140 credits |
What you can make with Veo 3.1
Realism and synced sound sell these six formats; start from the sample prompt on each.
Talking scenes and skits
Two people, one or two lines each, a place you can hear. Short comedy and drama are where Veo 3.1 is at its best.
Try this promptA customer at a hardware store holds up a tiny screw and asks "Do you have this in a bigger size?" The clerk stares and says "That is the bigger size." Store hum, a beep.
Ads with a voice-over line
A product moment and a spoken line in the same clip, with no separate recording.
Try this promptClose-up of hands opening a box of chocolates on a wooden table, a warm voice says "Made by hand, one batch a day." Paper rustling, soft piano.
Realistic B-roll
City streets, kitchens, offices, nature. Clips that cut in next to real footage.
Try this promptEarly morning, a fisherman pushes a small wooden boat into calm water, mist over the lake, birds calling, oars knocking against the hull.
Bringing a photo to life
An old family picture, a product photo or a painting as the first frame, with motion and sound added.
Try this promptAnimate this old photo: the couple turns toward each other and laughs, the wind moves the curtains behind them. A gramophone playing faintly.
Short film openings
An establishing shot with weather, mood and one line that sets up the story.
Try this promptNight, a lone truck stop in the desert, neon sign flickering, a woman steps out of her car, looks around and says "He said he would be here." Crickets, a buzzing sign.
Explainers and tutorials
A presenter in a real-looking place saying one clear sentence to camera.
Try this promptA friendly woman in a sunny garden kneels by a tomato plant and says to camera "Water at the roots, never the leaves." Birds, a light breeze.
Veo 3.1 vs Veo 3.1 Fast, Seedance 2.0 and Kling 4.0
The models people weigh against Veo 3.1 most often. Pick by what the shot needs: speech, detail, motion or a lower price.
| Veo 3.1 | Veo 3.1 Fast | Seedance 2.0 | Kling 4.0 | |
|---|---|---|---|---|
| Best for | Realistic scenes and dialogue | Drafts and volume | Longer multi-shot clips | Long takes and fast action |
| Resolution | 720p or 1080p | 720p or 1080p | 480p, 720p or 1080p | 720p or 1080p |
| Clip length | 8 s | 8 s | 5, 10 or 15 s | 5, 10 or 15 s |
| Sound | Yes, generated with the video | Yes, generated with the video | Yes, generated with the video | Yes, generated with the video |
| Ways to start | Text/Image, Frames | Text/Image, Reference, Frames | Text/Image, Reference, Frames | Text/Image, Reference, Frames |
| Credits, shortest clip at the lowest size | 200 | 28 | 66 | 140 |
Veo 3.1 prompt guide
A Veo 3.1 prompt reads best in this order: shot and camera, subject, action, place and light, then the words and the sounds.
- Start with the shot. "Handheld close-up", "wide establishing shot", "slow dolly in". Veo takes film language literally.
- Put speech in quotation marks and say who speaks: He says "…". Keep each line short enough to say in a few seconds.
- Fit the clip. 8 seconds holds one or two lines and one action. More than that gets rushed.
- Name two or three sounds: rain on glass, a kettle, traffic. Ambience makes the realism.
- For photo to video, describe only the motion. The picture already shows what things look like; tell Veo what changes.
- Say "no subtitles" if captions appear on a clip with speech.
Veo 3.1 image to video tips
Photo to video is one of the things Google says improved most in Veo 3.1. A few habits get the most out of it.
- Use the photo as the opening frame. Veo keeps it; the prompt only has to say what happens next and what it sounds like.
- Give people a line. A portrait that speaks one short sentence in quotation marks is the classic Veo clip.
- Keep motion believable for the photo. Small gestures, a head turn, wind in hair and clothes, a slow camera move.
- Match the ratio of the photo to the clip, vertical for Reels and Shorts, wide for YouTube.
- Use Frames to Video when you also know the last picture: a before and after, a reveal, a place changing with the light.
- Old or painted photos work too. Say "animate this photo" and describe the mood of the sound: a gramophone, a sea breeze, a crowd.
Veo 3.1 vs Veo 3.1 Fast
Both make 8 second clips at 720p or 1080p with sound and take the same prompts. Fast starts at 28 credits against 200 for Veo 3.1.
Use Fast to try lines, angles and timing; use Veo 3.1 for the take you will publish, where faces, hands and fine textures show. Fast is also the one with Reference to Video here.
What Google announced
Google introduced Veo 3.1 on October 15, 2025. Its post describes richer generated audio, now also in Ingredients to Video, Frames to Video and Extend; Ingredients to Video, which builds a clip from several reference images for characters, objects and style; Frames to Video between a start and an end image; Extend, which continues a clip from its last second to a minute or more; and stronger prompt adherence, better image to video and more realistic textures. It launched in Flow, the Gemini API, Vertex AI and the Gemini app.
On Genvima, each Veo 3.1 clip is 8 seconds at 720p or 1080p, from text, a photo or two frames. Extend and the Flow editing tools are part of Google’s own apps; join clips in an editor for longer pieces.
Limits worth knowing
- Length. Every clip is 8 seconds. Plan one moment per clip and cut several together.
- Long speeches get rushed or cut. One or two short lines per clip.
- Written words in the shot can still come out wrong even on Veo; overlay titles afterwards.
- Real identities. Google’s safety filters block some real people and some photos of real people; a blocked render is refunded in full.
Veo 3.1 vs Seedance 2.0 and Kling 4.0
Seedance 2.0 makes longer clips (5, 10 or 15 seconds) with sound and cuts between shots, from 66 credits. Kling 4.0 is the pick for long camera moves and fast action. Veo 3.1 wins on realistic people speaking in a real-looking place.
For close-up detail, also try Seedance 2.5. Credits work the same on every model, and pricing shows how many a plan or pack adds.
FAQ
What is Veo 3.1?
Veo 3.1 is Google’s AI video model, introduced on October 15, 2025. It generates realistic video with dialogue, sound effects and ambience in the same pass, and works from text, an image, reference images or a start and an end frame.
How can I use Veo 3.1 online?
Choose Veo 3.1 or Veo 3.1 Fast in the menu above and generate from text, a photo or two frames. You pay per clip in credits after a Google sign-in, instead of buying Flow, Gemini or Vertex access.
Is Veo 3.1 free?
No. Veo 3.1 is paid with credits: you sign in with Google, buy a plan or a credit pack, and every clip shows its price on the Generate button before you spend anything. To try an idea first, the free Basic engine in the same menu gives you one free video and three free images a day, with no credit card.
How long are Veo 3.1 videos?
On Genvima every Veo 3.1 and Veo 3.1 Fast clip is 8 seconds, at 720p or 1080p. Google’s Extend feature, which makes longer videos from several clips, is part of its own apps.
Does Veo 3.1 generate audio?
Yes. Dialogue, sound effects and ambience are generated with the picture. Put spoken lines in quotation marks in the prompt.
What is the difference between Veo 3.1 and Veo 3.1 Fast?
Fast costs less (from 28 against 200 credits) and takes the same prompts. Veo 3.1 is the one for the final take, where faces and textures show.
What is Ingredients to Video in Veo 3.1?
Google’s name for building a clip from several reference images: characters, objects and a style. On Genvima the same job is Reference to Video, on Veo 3.1 Fast.
Frames to Video or Ingredients: which should I use?
Frames to Video when you know exactly how the shot starts and ends. Ingredients (Reference to Video) when you know who and what is in it but want Veo to choose the shot.
How much does Veo 3.1 cost?
Veo 3.1 starts at 200 credits and is 200 at 1080p; Veo 3.1 Fast starts at 28. Credits come with a plan or a credit pack.
Is Veo 3.1 free?
No. Each Veo clip costs credits, starting at 28 on Veo 3.1 Fast; the full price appears on the button before anything is charged.
Veo 3.1 vs Seedance 2.0: which is better?
Veo 3.1 for realistic people speaking in 8 second clips; [Seedance 2.0](/video/seedance-2-0) for longer multi-shot clips at a lower price.
What happens if a Veo clip fails?
Nothing is lost: when Veo errors or blocks a prompt, the credits for that clip are returned automatically.
Keep exploring
- Seedance 2.0ByteDance’s video model with sound: multi-shot clips up to 15 seconds, from text, a photo or reference images.Open
- Seedance 2.5ByteDance’s newest video model: finer skin, fabric and light, multi-shot stories and sound in one pass.Open
- Kling 4.0Kuaishou’s newest video model for long, steady camera moves and fast action that stays smooth.Open
- Tiny Desk AIYou on lead vocals in the viral desk concert, your friends as the band.Open
- AI Video MakerEvery way to make a video and all 13 video models.Open
- PricingPlans and credit packs. One balance pays for every video and image model.Open
Try Veo 3.1 now
Write the idea, pick the model, and see the price before you generate.
Back to the prompt box
Veo 3.1
Seedance 2.5
Kling 4.0