Gemini Omni 1.1
Google · AI videoGemini Omni 1.1 Flash is Google's high-performance multimodal model for fast video generation and editing with natively synchronized audio. Unlike Veo, it runs on Google's Interactions API, which processes text, image, audio, and video together for more cohesive, controllable output.
On VidMachine it runs via Replicate at a fixed 720p for 30 credits per second, with clips steered to about 3–10 seconds.
Key features and benefits
Native audio
Every clip includes a synchronized audio track—dialogue, sound effects, and ambient sound—generated with the video.
Text-to-video and image-to-video
Generate from a detailed prompt alone, or animate a starting frame. Optional end-frame keyframe interpolation creates smooth transitions between two stills.
Prompt-timed control
Describe timing in the prompt (for example “after 3 seconds…” or [0-3s] / [3-6s] ranges). VidMachine also appends an explicit duration target for scene length.
Technical specifications
ProviderGoogle (Gemini Omni 1.1 Flash)
Resolution on VidMachine720p (fixed)
Aspect ratio16:9 or 9:16
DurationAbout 3–10 seconds
Pricing on VidMachine30 credits per second
Use cases and applications
Fast cinematic shorts with native audio when you want Google multimodal quality without Veo pricing.
Animate a scene start frame, or interpolate between start and end frames for orbits, zooms, and seamless loops.
Why this model
Choose Gemini Omni 1.1 when you want Google-native audio video with strong prompt control at a fixed 720p rate.
How VidMachine uses it
VidMachine calls google/gemini-omni-1.1 on Replicate at 720p. Scene pipelines pass the starting frame as image and an optional last frame. Aspect ratio follows the project setting.
Billing is 30 credits per second of output video. Requested length is clamped to 3–10 seconds and steered in the prompt; final billing uses the probed clip duration.
What you should know
What resolution does VidMachine use?
Always 720p. Higher tiers like 1080p and 4K are not selectable.
Does Gemini Omni generate audio?
Yes. Every generated video includes a synchronized native audio track.
How long can a clip be?
About 3–10 seconds. VidMachine clamps scene targets into that range and bills by actual output length.