Veo 3.1 Fast
Google's Veo 3.1 Fast turns a text prompt, optionally with a first frame or a first and last frame, into a 4 to 8 second video with dialogue, sound effects and ambient audio, at 720p, 1080p or 4K.
Overview
Veo 3.1 Fast is the lighter variant of Google's Veo 3.1 video model. From a text prompt it returns a short clip whose dialogue, sound effects and ambient audio follow the action. It can also start from a still image, or generate the motion between a first frame and a last frame you supply.
Google's launch post for Veo 3.1 highlights richer native audio, more narrative control and better prompt adherence when a video starts from an image. On Wiro the model produces clips of 4, 6 or 8 seconds in landscape (16:9) or portrait (9:16), at 720p, 1080p or 4K.
What you can build
- Vertical social clips and short ads with spoken lines and sound cues written into the prompt.
- Photo and product animation: upload a still as the first frame and describe the motion, as in the duckling sample on this page.
- Transitions and before/after shots built from a first and a last frame.
- Storyboards and concept tests: compare prompt directions in 720p clips, then run the prompt you settle on at 1080p or 4K.
Inputs
- A text prompt describing the shot, the camera movement and the sound. The sample prompts here put the audio in the same sentence as the picture: a stadium crowd chanting, a spray can hissing, dozens of clocks ticking.
- An optional first-frame image (JPG, PNG, WebP, GIF or HEIC).
- An optional last-frame image, used together with a first frame for interpolation.
- Aspect ratio: 16:9, 9:16, or match the input image (the default). When matching, the clip takes whichever of the two ratios is closer to your image. With no image it is 16:9.
- Resolution: 720p (default), 1080p or 4K.
- Duration: 4 (default), 6 or 8 seconds.
- A negative prompt listing what to keep out of the video, and an optional seed.
Outputs
One MP4 video at 24 frames per second with a generated soundtrack. Audio is always on for this model. Google adds a SynthID digital watermark to every Veo video.
Recommended settings
- Write prompts in English. Google fully supports English and has not evaluated other languages.
- Cover the subject, the action, the style and the camera position or move, then the sound: effects, ambient noise and dialogue. Put dialogue in quotes so the model treats it as speech.
- Google's rules for Veo 3.1 tie 1080p and 4K to 8-second clips. A 4 or 6 second clip is a 720p option.
- A fixed seed makes reruns closer to each other, but Google does not promise identical output.
Limitations
- One request gives one clip of at most 8 seconds, so longer pieces are cut together from several clips.
- When a video starts from a first frame, or a first and last frame, Google only allows adults in it, and an input image that shows a child can be refused.
- Google sometimes blocks a video because of its safety filters or a problem processing the audio. The request then returns no video.
Safety & compliance
Google runs safety filters on prompts, uploaded images and generated videos. The SynthID watermark can be checked on Google's SynthID platform to confirm that a clip is AI-generated. Use only images you have the rights to, and label synthetic footage where your platform or local law requires it.
Example prompts
Great starting points for veo3.1-fast.
API quick start
Run veo3.1-fast with a single API call.
{
"prompt": "A sweeping, elevated wide shot captures t…",
"durationSeconds": 4,
"inputImage": "https://your-cdn.com/input.png",
"inputImage2": "https://your-cdn.com/input.png"
}curl -X POST "https://api.wiro.ai/v1/Run/google/veo3-1-fast" \
-H "Content-Type: application/json" \
-H "x-api-key: YOUR_WIRO_API_KEY" \
--data-binary @- <<'JSON'
{
"prompt": "A sweeping, elevated wide shot captures t…",
"durationSeconds": 4,
"inputImage": "https://your-cdn.com/input.png",
"inputImage2": "https://your-cdn.com/input.png"
}
JSONOther models you might like
Explore similar video generation models.