Google News

WebMCP: Chrome's Early Preview of an Agent-Ready Web Standard

RSS

On 10 February 2026, Google’s Chrome team published an early preview of WebMCP (Web Model Context Protocol), a proposed web standard that lets websites declare callable tools for in-browser AI agents. The announcement is early-stage infrastructure, not an immediate ranking factor, but it marks the first formal attempt to standardise how AI agents interact with the web at the browser level.

What WebMCP is

MCP (Model Context Protocol) was introduced by Anthropic in late 2024 as a standard for connecting AI models to external tools and data sources. It has since been adopted broadly across the AI industry. WebMCP is an adaptation of that protocol for the web, allowing a site to declare its own tools in a structured format that a browser’s AI agent can discover and call.

Where a conventional web page communicates with users through HTML that humans read, a WebMCP-enabled page also communicates with AI agents through a structured interface that machines can invoke.

How it works

A site implementing WebMCP defines a set of tools in a standardised format, published at a known endpoint. An in-browser AI agent, such as Gemini in Chrome, can discover these tools and call them on the user’s behalf.

The practical difference is efficiency. An AI agent trying to search an e-commerce site without WebMCP would need to simulate user behaviour: navigate to a search box, type a query, wait for results to load, scroll through pagination, and extract information from rendered HTML. With WebMCP, the same agent calls a searchProducts function, passes structured parameters, and receives structured JSON results in a single operation.

Google’s early preview documentation gives this as an example of how a single WebMCP tool call can replace what might otherwise require dozens of browser interactions.

Difference from structured data

WebMCP and structured data (schema markup) address different questions. Structured data says what a page contains: this is a Product, it has a name, price, and review rating. WebMCP says what can be done on a page: here is a checkAvailability function, here is an addToCart function, here is a bookAppointment function.

The two are complementary. Structured data improves how AI systems understand and index content. WebMCP enables AI agents to act on content on behalf of users. Sites implementing both would give AI agents both the context to understand what they are looking at and the tools to take action on it.

Current state and timeline

WebMCP is available in Chrome 146 Canary behind the “WebMCP for testing” flag at chrome://flags. It is in early preview for developers who want to experiment with the standard, not a feature available to general users.

Full Chrome and Edge support is expected mid-to-late 2026, pending the standard maturing and adoption increasing.

What this means

For most sites, WebMCP is not a near-term priority. The surrounding infrastructure is early-stage: few AI agents currently support WebMCP, browser support is limited to Canary, and the standard itself may evolve before it reaches stable release.

The value in understanding WebMCP now is strategic rather than tactical. As AI agents become more capable and more widely used, the sites that have invested in machine-readable interfaces will be better positioned than those that have not. The pattern is similar to the adoption curve of structured data: early implementers gained a lead that compounded over time.

Sites in categories where agentic tasks are most likely, e-commerce, travel, appointments, and local services, have the most to gain from early experimentation. For informational and editorial sites, the implications are further out.

Sources

More news