A Google video model that does everything in one place: shoot a clip from a description or a photo, edit a finished video with words, continue it and upscale to 4K. Version 1.2 raised the clip length from 10 to 15 seconds, and sound is generated right away.
| Developer | |
|---|---|
| Modes | text, photo, references, editing, extension, upscale |
| Length | 4, 6, 8, 10 or 15 seconds |
| Quality | 360p, 720p, 1080p, 4K |
| Ratios | 9:16, 16:9 |
| Sound | always generated |
| Editing | clips up to 15 seconds |
| Extension | +4–10 seconds per step |
| Quality | 4 seconds | 8 seconds | 15 seconds |
|---|---|---|---|
| 360p | $0.15 | $0.29 | $0.54 |
| 720p | $0.43 | $0.86 | $1.60 |
| 1080p | $0.64 | $1.28 | $2.40 |
| 4K | $1.28 | $2.56 | $4.80 |
The same price for a clip from text and from a photo. Sound is included.
Speech, noises and ambience are generated with the picture. No separate voiceover to pay for, and sound cannot be switched off.
360p is a third of the price of 720p: try ideas at $0.15 a clip and render only the good one in 1080p or 4K.
Upload your clip up to 15 seconds and say what to change: clothes, background, time of day, the object in a hand. The motion and cut stay.
Each step adds 4–10 seconds at the end, using the last 10 seconds as context. You pay only for the new seconds.
One clip is up to 15 seconds. Extension grows it in steps of 4–10 seconds.
9:16 and 16:9. The model has no square.
Yes, in Photo to video. In that mode the length is capped at 10 seconds by the model itself.
From 720p and up, yes. The model refuses 480p sources, so shoot at least in 720p first.