Gemini 3.5 Flash Gets Native Computer Use: Google's AI Agent Navigates Any Screen Autonomously
View original source →On June 24-25, 2026, Google announced that computer use is now a built-in native tool inside Gemini 3.5 Flash—the same production model already used by enterprise developers for function calling, Search grounding, and Maps integration.
Key points:
• Computer use was previously accessible only through a standalone Gemini 2.5 Computer Use model requiring separate endpoints; native integration eliminates this routing complexity
• Gemini 3.5 Flash with computer use scores 78.4 on OSWorld-Verified benchmark—matching GPT-5.5's 78.7 (0.3 point difference) at roughly one-third the per-token cost
• The practical capability: an agent can receive a high-level task ('Book the cheapest flight from Newark to Toronto on July 18'), navigate airline sites, compare prices, complete booking, and create calendar events—all without human intervention
• Enterprise deployments can scope screen access to specific application windows or browser tabs, providing sandboxed execution environments
• Integration with Search and Maps is significant: a computer use agent that can also search the web and query mapping data represents a more capable autonomous workflow executor than screen-only agents
• Available through the Gemini API and Gemini Enterprise Agent Platform
Why It Matters: Enterprise automation can now reach workflows that API-based automation cannot—specifically legacy applications with no API, complex web applications, and desktop software never designed for machine interaction. IT teams with processes stuck behind UI-only interfaces now have a viable AI automation pathway at Flash pricing.