Vercel MCP is Vercel's official remote MCP server with OAuth, hosted at https://mcp.vercel.com. It lets AI assistants search Vercel documentation, manage teams, projects and deployments, and query Web Analytics. It is in Beta on all plans and works only with AI clients reviewed and approved by Vercel.
What it does
Where Vercel MCP runs and what it adds, from its documentation
Search and navigate Vercel documentation (public tools, no auth needed)
Manage projects and deployments
Analyze deployment logs
Query visitors, page views and custom events from Web Analytics
Implements MCP Authorization and Streamable HTTP specifications
When to use it
Good fit
You want your coding agent to inspect deployments and logs without leaving the editor
You want to debug a failed Vercel deployment with an AI assistant
You want one command to set Vercel up across many MCP clients
Not for
Clients not on Vercel's approved list (only reviewed clients can connect)
Workflows where the agent also reads untrusted content without human confirmation
How to install Vercel MCP
You need: A Vercel account (OAuth login) for authenticated tools; An MCP client approved by Vercel; Claude custom connectors need Pro, Max, Team or Enterprise plan
Run: npx -y add-mcp https://mcp.vercel.com -g (configures all detected clients)
Or in Claude Code: claude mcp add --transport http vercel https://mcp.vercel.com
Or in Codex CLI: codex mcp add vercel --url https://mcp.vercel.com
Start the client and authenticate with Vercel (in Claude Code type /mcp)
Not stated by the maintainers: No public source repository for the MCP server found; repo_url points to the docs page; License; Tool list not fetched (separate tools page); Last release date.