The announcement, published on May 19, 2026 by the Chrome for Developers team, outlines a broad set of additions spanning developer tools, web performance capabilities, and end-user AI features powered by Gemini.
Google Is Teaching AI Agents How to Use Websites, Properly
At the centre of the developer-facing announcements is WebMCP, a proposed open web standard that allows website owners to expose structured tools, including JavaScript functions and HTML forms, directly to browser-based AI agents. Rather than having an agent simulate human clicks through a website, WebMCP enables it to call machine-friendly functions to complete tasks with greater reliability and speed. An experimental origin trial opens in Chrome 149, and Google confirmed that Gemini in Chrome will soon support WebMCP APIs. A number of global brands, including Expedia, Shopify, Instacart, Target, and Etsy, are already noted as early experimenters.Alongside WebMCP, Google is introducing Modern Web Guidance, a curated set of expert-vetted instructions for AI coding agents, covering over 100 use cases. The tool, available in early preview, integrates with Google's Baseline standard to ensure generated web code is accessible, performant, and secure. It can be installed via a single click in Google Antigravity, through npx, or as an agent extension.
Chrome DevTools for agents is also now available, providing AI coding tools with direct access to console logs, network traffic, and accessibility trees for automated debugging. LY Corporation, cited in the official release, used the tool to reduce manual performance analysis by 96–98%.
Chrome's Built-in AI Is Getting Smarter and Cheaper to Run
Chrome's built-in AI capabilities are expanding. The Prompt API is now stable in Chrome 148, using Gemini Nano with multimodal inputs and structured JSON output. A new ultra-efficient model, Gemma 197M, has been introduced to power task-specific APIs such as the Summarizer, enabling AI features to scale across a wider range of devices without server costs. Travel platform Trip.com is cited as using built-in AI to generate personalised travel summaries entirely on-device.Websites Are About to Look and Feel More Like Apps
On the web UI front, Google announced the HTML-in-Canvas API, now available as an origin trial, which allows real DOM elements to be embedded directly into a canvas with WebGL and WebGPU support, enabling 3D, searchable, and accessible web environments. This is paired with element-scoped view transitions, available in Chrome 147, to deliver fluid, app-like animations without blocking page interactivity.The Soft Navigations API, arriving in an upcoming Chrome release, will extend Core Web Vitals measurement to Single Page Applications (SPAs), a long-standing gap in web performance monitoring.
Gemini Is Coming to Chrome on Android This June
Gemini in Chrome, already available on desktop and iOS, is coming to Android in late June 2026. The Android version includes article summarisation, integration with Google apps like Calendar, Keep, and Gmail, and a contextual feature called Personal Intelligence that can tailor responses using data from connected apps like Gmail and Google Photos.Auto browse, a task automation feature that can handle multi-step digital tasks such as booking parking or comparing products, is also arriving on Android. On desktop, auto browse will be integrated with Gemini Spark in the coming months.
I truly appreciate you spending your valuable time here. To help make this blog the best it can be, I would love your feedback on this post. Let me know in the comments: How could this article be better? Was it clear? Did it have the right amount of detail? Did you notice any errors?
If you found any of the articles helpful, please consider sharing it.