WebMCP in 2026: Which Browsers Support navigator.modelContext? (Complete Compatibility Status)

#webmcp #browsers #ai #javascript

Chrome 146 Canary shipped navigator.modelContext last week. That makes Chrome the first browser with native WebMCP support - giving AI agents structured access to web page content without scraping, without Puppeteer, without fragile CSS selectors.

But Chrome isn't the whole web. If you're building a WebMCP-compatible site today, you need to know where every browser stands. Here's the complete picture as of March 2026.

Browser Compatibility Table

Browser	WebMCP Status	`navigator.modelContext`	Version	Notes
Chrome	Supported	Yes	146 Canary	First implementation. Behind flag in stable.
Firefox	In Progress	No	-	Mozilla is in the W3C WebMCP working group. No public timeline.
Safari	In Progress	No	-	Apple joined the W3C working group. WebKit tracking bug filed.
Edge	In Progress	No	-	Microsoft participating in W3C spec. Likely ships shortly after Chrome stable (shared Chromium base).
Brave	Not Started	No	-	No public commitment. Chromium-based, so could adopt Chrome's implementation.
Arc	Not Started	No	-	Chromium-based. No public WebMCP work.

Bottom line: Chrome 146 is the only browser with a working implementation right now. Edge will almost certainly follow quickly since it shares Chrome's engine. Firefox and Safari are engaged in the spec process but haven't committed to timelines.

What This Means for Developers

If you're building an agent-ready website today, you can't assume WebMCP is available. You need graceful degradation - serve structured agent data when navigator.modelContext exists, fall back gracefully when it doesn't.

That's exactly what webmcp-sdk handles.

Cross-Browser Implementation with webmcp-sdk

import { WebMCPProvider, isWebMCPSupported } from 'webmcp-sdk';

// Check for WebMCP support before using it
if (isWebMCPSupported()) {
  // Full WebMCP - agent gets structured context
  const provider = new WebMCPProvider({
    actions: ['search', 'add-to-cart', 'checkout'],
    schema: productCatalogSchema
  });
  provider.expose();
} else {
  // Fallback: expose data via meta tags + JSON-LD
  // Agents using standard web parsing still get structured data
  injectStructuredFallback(productCatalogSchema);
}

The SDK does three things:

1. Feature detection. isWebMCPSupported() checks for navigator.modelContext at runtime. No user-agent sniffing. No browser lists to maintain.

2. Structured data exposure. When WebMCP is available, the SDK exposes your page's actions, data schemas, and interactive elements through the native API.

3. Graceful fallback. When WebMCP isn't available, the SDK injects structured data via JSON-LD and semantic HTML. Agents using traditional web access still get clean, typed data - just through a different channel.

The W3C Working Group

The WebMCP specification is being developed through the W3C with participation from Google, Microsoft, Mozilla, and Apple. That's a strong signal. When all four major browser vendors are at the table, the spec usually ships across browsers within 12-18 months of the first implementation.

Chrome 146 is the reference implementation. The spec is still evolving, but the core API surface - navigator.modelContext for reading page context, ModelContextProvider for exposing agent-compatible data - is stable enough to build on.

Should You Build for WebMCP Today?

Yes. Here's why:

Chrome holds roughly 65% of browser market share. Chrome 146 hitting stable means the majority of your users' browsers will support WebMCP within months. And with the webmcp-sdk fallback pattern, you're not breaking anything for Firefox or Safari users - they just get the non-WebMCP version.

Sites that add WebMCP support now will be indexed and ranked by AI agents before their competitors. It's the same dynamic as mobile-responsive design in 2012 - early adopters got the SEO advantage.