Veo 3.1 on VisionX.
Google video with native audio, in three grades from the finished take to the fast draft.
What Veo 3.1 is on the roster
Veo 3.1 is the Google route on the roster, and it runs in three grades. Veo 3.1 is the full model; Veo 3.1 Fast is Google’s quicker, lower-cost grade; Veo 3.1 Lite is the draft grade for animatics and volume. All three render sound with the picture, and any shot can be switched to silent.
A shot can start from text, open on a frame you supply, or travel between a first and a last frame. Veo 3.1 and Fast also take up to three reference images and reach 4K. Every grade renders your Cast, prices in VX before the run, and sits one dropdown away from a Seedance or Kling take of the same board line.
Native audio, with an off switch
Dialogue, ambience and effects render in the same pass as the picture. When the cut will be scored later, switch audio off for that shot.
Open on a frame, or land on one
Image-to-video from a first frame, or a first and last frame the shot has to travel between. The output inherits the framing you hand it.
Three references for subject and style
Veo 3.1 and Fast take up to three reference images, so a Cast member, a product and a look can brief one shot. Reference shots render at eight seconds.
Up to 4K
720p and 1080p on every grade, and 4K on Veo 3.1 and Fast. 1080p and 4K shots run eight seconds.
Three grades, one board
Block a scene on Lite, iterate on Fast, and finish the keeper on Veo 3.1. The composer quotes each grade before you run it.
What Veo 3.1 actually does
Three grades of Google’s video model, and what each shot can start from.
Three grades of one model
Veo 3.1 for the finished take, Veo 3.1 Fast for iteration at a lower cost, and Veo 3.1 Lite for animatics and volume. The prompt that works on one works on the others.
Draft to finalStart from a frame
Hand it a still and the shot opens on it. A keyframe from the image generator, a Cast frame or a product plate becomes the first frame of the take.
Image to videoA first frame and a last frame
Give it both ends and Veo works out the move between them, so a shot can be told where to land instead of being re-rolled until it happens to.
First and last frameThree references, one shot
Veo 3.1 and Fast take up to three reference images for subject, product and style. A reference shot renders at eight seconds.
Reference imagesSound in the same pass, or none
Dialogue, ambience and effects are generated with the picture. Switch audio off for a shot that will be scored in the edit.
Native audio720p to 4K
Every grade renders 720p and 1080p; Veo 3.1 and Fast also render 4K. Shots at 1080p and 4K run eight seconds. At 720p a text, single-frame or first-and-last-frame shot can run four, six or eight; reference-image shots are eight seconds.
ResolutionEvery variant, exactly as the studio runs it
These rows are read live from the engine registry and the pricing engine. What you see here is what the composer quotes.
Self-serve rates run as low as $0.085 per VX at scale. The composer quotes the exact VX for your shot, references included, before you run it.
How to write for Veo 3.1
Veo reads a prompt as a shot description with a soundtrack attached. Say what the camera sees, then what the audience hears.
The founder from the first frame sits at her workbench, turning a prototype in her hands. Slow handheld push-in, shallow focus, late-afternoon window light. Audio: room tone, a soft tick of the clock, and her line, "We built the first one in this room."
Write the sound into the prompt
Audio renders with the picture, so a line of dialogue, the ambience and any effect belong in the prompt. Quote dialogue exactly as it should be spoken.
Let the frames carry the composition
When a first or last frame is attached, describe the motion and the performance rather than re-describing what the frame already shows.
Pick the rung before the length
1080p and 4K shots are eight seconds, and so is a reference shot. For a four or six second beat, render a text, single-frame or first-and-last-frame shot at 720p.
Compose at the delivery ratio
Veo renders landscape 16:9 and vertical 9:16. When a first frame is attached, compose it at the ratio you want delivered so the approved framing survives.
What teams route to Veo 3.1
What producers reach for Veo to cover.
Talking-head and interview coverage
Naturalistic performance with sound already in the pass.
Documentary-flavoured b-roll
Handheld realism that reads as captured rather than generated.
Animating approved key art
A signed-off still becomes the first frame, so the film opens exactly where the campaign stills left off.
Animatics on Lite
Block a whole sequence on the draft grade, then re-run the keepers on Veo 3.1.
Same Cast, same wallet, different look
Route a board line to Veo 3.1, then run the identical shot on another family and compare takes. Identity lives in your Cast, not in any one model, so switching engines never strands the campaign.