Google introduced two new artificial intelligence models — Gemini Omni and Gemini 3.5 Flash — during its I/O 2026 developer conference on Wednesday. The models highlight the company's progress in three key areas: video creation, agentic workflows, and coding.
What the models can do with video
The advances in video creation mean the models can generate and edit video clips based on text prompts. Google demonstrated this during the keynote, showing the AI creating a short scene from a description and then modifying it with follow-up instructions. The models understand visual context better than
