home / mcp / chrome devtools mcp server
Provides programmatic access to Chrome DevTools via MCP for automation, debugging, and performance insights.
Configuration
View docs{
"mcpServers": {
"chromedevtools-chrome-devtools-mcp": {
"command": "npx",
"args": [
"-y",
"chrome-devtools-mcp@latest",
"--browser-url=http://127.0.0.1:9222",
"-y"
],
"env": {
"CHROME_DEVTOOLS_MCP_NO_USAGE_STATISTICS": "true"
}
}
}
}Chrome DevTools MCP enables your AI coding assistant to control and inspect a live Chrome browser through a dedicated MCP server. It unlocks reliable automation, in-depth debugging, and performance analysis by exposing Chrome DevTools capabilities to your MCP client.
You connect an MCP client to the Chrome DevTools MCP server and start a Chrome instance that the MCP can control. Use this to perform automated browsing tasks, analyze network activity, take screenshots, run performance traces, and inspect console messages. Start by adding the server configuration to your MCP client, then issue prompts that drive browser actions or request insights. The MCP server will launch or connect to a Chrome instance as needed and return results back to your MCP client.
Prerequisites: you need Node.js and npm installed on your machine, and a Chrome browser installed (current stable or newer). Follow these steps to set up the Chrome DevTools MCP server for use with your MCP client.
Configuration and usage notes: You can run the server in standard mode or in slim mode for basic browser tasks. To connect to a running Chrome instance, you can enable automatic connection (Chrome 144+) or connect manually via a remote debugging port. Use the provided command examples to start the MCP server via your MCP client. Usage statistics are enabled by default and can be opt-out with a specific flag. Privacy considerations and data collection practices follow Google policies.
The MCP server exposes the browser contents to MCP clients, enabling inspection, debugging, and modification. Avoid sharing sensitive data through MCP clients. If you enable performance traces that use external data sources, be aware of data sharing practices and disable features if needed.
If the browser does not start automatically, ensure you are issuing prompts that require a running browser. When connecting to a running Chrome instance, verify remote debugging is enabled and the browser URL or WebSocket endpoint is correctly configured in your MCP client. For issues with remote debugging, consult relevant Chrome DevTools documentation and ensure port accessibility for your environment.
Use cases include performance tracing of web pages, analyzing network traffic, capturing screenshots for bug reports, and extracting console messages for debugging. You can automate repetitive browser tasks and generate actionable insights from traces and logs.
Automates clicks, drags, form filling, and interactions with page elements.
Controls page navigation, window management, and waiting for conditions.
Emulates device characteristics and resizes pages.
Analyzes performance traces, starts and stops traces, and analyzes memory usage.
Inspects and lists network requests and responses.
Evaluates scripts, retrieves console messages, runs audits, and takes screenshots or snapshots.