← Back to MCP servers

Vercel

Vercel MCP Server

Manage Vercel deployments, projects, and logs from your AI coding assistant.

Token requiredGet credentials →

Last reviewed Mar 2, 2026

Install

claude mcp add vercel --transport stdio -- npx -y vercel-mcp-server
Scope: project

Environment variables required:

VERCEL_API_TOKEN=your-vercel-api-token

Available tools

ToolDescription
list_deploymentsList recent deployments with status and URLs
get_deploymentGet details about a specific deployment
list_projectsList all projects in your Vercel account
get_logsRetrieve build or runtime logs for a deployment
create_deploymentTrigger a new deployment

What it does

The Vercel MCP Server connects your AI coding assistant to Vercel's deployment platform, enabling it to monitor deployments, inspect build logs, manage projects, and trigger new deployments. This keeps your entire deployment workflow accessible from your editor.

Having deployment context available to your assistant is particularly valuable when debugging production issues. It can check recent deployment statuses, review build logs for errors, and help you understand what changed between deployments -- all without switching to the Vercel dashboard.

Use cases

  • Monitor deployment statuses to verify your latest push deployed successfully
  • Review build logs when deployments fail to quickly identify and fix issues
  • List projects and their deployment URLs for quick reference during development
  • Trigger new deployments from your editor as part of your workflow
  • Compare deployment details across environments (preview, production)

Getting started

  1. Create a Vercel API Token at vercel.com/account/tokens. Choose a descriptive name and select the scope (personal account or team).
  2. Set the VERCEL_API_TOKEN environment variable to your token value.
  3. The token scope determines which projects and teams the assistant can access. For team projects, ensure the token has team-level access.
  4. Deployment operations require appropriate permissions. Read-only tokens are recommended if you only need to monitor deployments and view logs.