Gemini Omni Flash
by Google Soon

Gemini Omni Flash is a Google video-first media generation model for multimodal storytelling and world simulation. It accepts text, image, video, and audio guidance to generate video with native audio.

Gemini Omni is marked Soon on Krea. This landing page is prepared for the planned model id gemini-omni as availability and generation settings land in the Krea Video Tool.

Gemini Omni Flash works with detailed prompts plus image, video, and audio inputs. That makes it useful for briefs that need visual references, motion direction, and sound cues together.

Yes. Gemini Omni Flash can generate video with native audio, including speech and sound effects, and supports video-to-audio generation among its media-to-media capabilities.

Conversational editing means refining generated media with natural-language instructions. Gemini Omni is designed to support precise edits through conversation, especially inside Flow workflows.

Yes. Gemini Omni Flash outputs include SynthID watermarking and C2PA content credentials to help identify generated media.

When Gemini Omni is available on Krea, open the Video Tool, choose Gemini Omni, add your prompt and supported references, then generate from the same Krea workflow you use for other video models.

Gemini Omni is aimed at filmmakers, storytellers, marketers, designers, and creative teams that want one model to reason across prompts, references, video, and sound for richer AI video concepts.