Use Gemini Omni Flash as a Gemini AI video generator for text prompts, image references, multi-image references, and single-video edits in one workflow.
Create 720p AI videos from a prompt, one image, multiple reference images, or an input video edit brief. Gemini Omni Flash is built for fast multimodal video experiments with simple controls.
Write a compact brief that names the source media, the intended motion or edit, and the publishing format.
Create a 5-second rainy city night scene with a person in a long coat walking across neon reflections. Slow camera push, soft rim light, fine rain mist, cinematic short-form style.
Start with the smallest setting that proves the idea, then increase duration only when the motion needs more time.
Use text only for fast scene exploration and short-film ideation.
Upload one image for direct image-to-video, or multiple images when each visual reference has a different role.
Use one input video and describe the edit. Pricing follows the input clip duration.
This tool uses third-party model capabilities and is not affiliated with or endorsed by the original model providers unless explicitly stated.
Open Gemini Omni Flash, write a scene or edit instruction, optionally add one image, multiple reference images, or one video, choose 720p and the ratio, then generate and review the clip.
The current integration exposes 720p output with 16:9 and 9:16 ratios.
Use tags such as <IMAGE_REF_0> and <IMAGE_REF_1>. The prompt editor can insert these tokens for uploaded images.
Text-only requests use the published text-to-video per-second estimate. Requests with image or video input use the multimodal per-second estimate, and video edit pricing is based on input video duration.