Google AI video models
Google's video catalog covers two different starting points: creating a new shot with Veo and working with mixed inputs or existing footage with Gemini Omni. Choose the required input mode first, then compare turnaround, output options, and credits within that family.
5 models available in Starpop. Updated .
Compare Google video models
| Model / best use | Inputs | Output options | Estimated cost |
|---|---|---|---|
| GooglevideoGemini Omni Flash Source-video transformations | Text, Image, Video | 720p, 1080p4s, 6s, 8s, 10s | $0.15 USD / second1.5 credits / second |
| GooglevideoGemini Omni Flash 1.1 Source-video transformations | Text, Image, Video | 360p, 720p, 1080p, 4k3–10s | $0.03–$0.30 USD / second0.3–3 credits / second |
| GooglevideoVeo 3.1 Cinematic advertising shots with native audio | Text, Image | 720p, 1080p4s, 6s, 8s | $0.40 USD / second4 credits / second |
| GooglevideoVeo 3.1 Fast Rapid Veo prompt iteration | Text, Image | 720p, 1080p4s, 6s, 8s | $0.20 USD / second2 credits / second |
| GooglevideoVeo 3.1 Lite Low-cost storyboarding and drafts | Text, Image | 720p, 1080p4s, 6s, 8s | $0.10 USD / second1 credit / second |
Choosing a Google model
Creating a new shot with sound
Compare the Veo 3.1 routes for text-to-video and image animation with native audio. Keep subject action, camera movement, and sound in one timeline so they describe the same scene.
Choosing a Veo production route
Use Fast or Lite for prompt variations, and compare the approved result with standard Veo. Check the supported duration and resolution before planning a sequence.
Editing footage or combining references
Start with Gemini Omni when your brief depends on an existing video. Gemini Omni Flash 1.1 also supports frame pairs and mixed references. Explain which elements to preserve and which should change.