Google has published nine demonstration videos detailing the capabilities of its Gemini Omni and Gemini 3.5 Flash models following their announcement at Google I/O 2026. The new demonstrations highlight real-time conversational video editing, complex agentic coding, and automated search experience generation. Readers and developers can now evaluate how these models execute multi-step workflows and manipulate multimodal media.
Conversational Video Editing with Gemini Omni
Gemini Omni introduces native multimodal creation by accepting combined text, image, audio, and video inputs to generate or modify video content grounded in real-world knowledge. Through natural language prompts, users can perform multi-turn edits—such as changing lighting, adding objects, or shifting camera angles—while maintaining character consistency and physical realism across frames.
The model is rolling out globally to Google AI Plus, Pro, and Ultra subscribers via the Gemini app and Google Flow, alongside free access on YouTube Shorts and the YouTube Create App. API access for enterprise customers and developers is scheduled to follow in subsequent weeks.
Gemini 3.5 Flash and Agentic Workflows
Built to execute complex long-horizon tasks, Gemini 3.5 Flash combines frontier intelligence with rapid processing speeds tailored for software engineering and automated workflows. Powered by Google's Antigravity harness, 3.5 Flash coordinates collaborative subagents to categorize unstructured data, execute coding tasks, and generate custom interface designs.
Key operational capabilities highlighted across the video releases include:
- Generating different UX checkout flow designs in just 60 seconds on AI Studio.
- Deploying 24/7 background information agents to track topic updates and send web-linked summaries.
- Building on-the-fly generative UI visual tools, simulations, and custom trackers directly inside Google Search.
- Powering Gemini Spark, a personal agent for Workspace applications like Gmail, Docs, and Slides currently available to U.S. Google AI Ultra subscribers.
Key Access Restrictions and Caveats
While Gemini 3.5 Flash is generally available through Google Antigravity, Google AI Studio, Android Studio, and the global Gemini app, several advanced features remain restricted by subscription tier. Automated background information agents and persistent custom Search dashboards will initially launch only for Google AI Pro and Ultra subscribers, leaving base tier users without continuous agentic tracking. Additionally, full deployment across Workspace and search customization remains geographically restricted to U.S. subscribers during the initial rollout phase.