Google Introduces Gemini Omni Model for Video Generation and Editing

Google introduced Gemini Omni, a new family of multimodal AI models, at its I/O developer conference. The models can reason across text, images, audio, and video to generate and edit video output from combined inputs. ” The initial release focuses on video generation and text-based photo editing.

Omni processes multiple input types together rather than stitching them separately. It produces videos that reflect understanding of physics, culture, history, and science. Users can edit photos using plain text commands. The model also supports creation of videos with personal digital avatars after a dedicated onboarding process.

Omni Flash, the first model in the family, generates up to 10 seconds of video. It is rolling out today to the Gemini app, YouTube Shorts, and AI creative studio Flow. All videos created with the model include Google’s SynthID digital watermark. Google plans to release an API version in the coming weeks and a higher-performance Omni Pro model later.

Google DeepMind director of product management Nicole Brichtova said the release combines Gemini’s intelligence with media model rendering capabilities. Longer video durations are planned for future updates.