Google I/O 2026 introduced the 'Agentic Web' concept with sweeping Chrome updates across three areas. WebMCP, a proposed open standard, lets developers expose structured JavaScript functions and HTML forms to browser-based AI agents for complex task execution. Built-in AI via Gemini Nano and Gemma 197M brings on-device summarization and translation to Chrome 148's stable Prompt API, eliminating server token costs. New UI APIs include HTML-in-Canvas for 3D DOM integration, element-scoped view transitions in Chrome 147, and the Soft Navigations API in Chrome 150 for Core Web Vitals in SPAs. Gemini in Chrome is expanding to Android with Workspace integration, Auto Browse for task automation, and Gemini Spark for cross-tab agent actions on desktop.

4m read timeFrom sdtimes.com
Post cover image

Sort: