Announcing the Render MCP Server

Today, we’re launching the official Render MCP server into general availability! During our early access period, customers found the MCP server incredibly useful for debugging service issues, optimizing performance, and accelerating their product development cycles. Thanks for helping us make Render MCP ready for production workflows. We’re also introducing a powerful new capability: direct access to service metrics. You can now ask for CPU utilization, memory usage, and bandwidth consumption using natural language, adding a new dimension to performance monitoring and troubleshooting: The Render MCP server turns your favorite AI tools like Cursor and Claude Code into a powerful command center for your infrastructure.

Use cases

The Render MCP server unlocks a variety of workflows right from your terminal or IDE:

• Faster troubleshooting: Diagnose and resolve service and build issues the moment they happen, without context switching.
  • "Pull the 50 most recent error logs for my production API."
  • "Why isn't my site at example.onrender.com working?"
• Effortless provisioning: Describe the infrastructure you need in plain English, and your AI assistant will build it on Render.
  • "Create a new Postgres database named 'user-db' with 5 GB of storage."
  • "Deploy the example Flask app from the following GitHub URL"
• Intuitive data fetching: Query your databases directly from your editor without writing a single line of SQL.
  • "How many users signed up in the last 7 days?"
  • "What were our top 5 most purchased items last month?"
• Rapid performance analysis: Investigate and optimize service performance by querying live metrics.
  • "What was the peak CPU usage for my web service in the last 24 hours?"
  • "My site feels slow. Compare the CPU and memory metrics from before and after the last deployment for 'my-production-api'."
  • "My API service is on the Pro plan. Based on its peak resource utilization over the last 30 days, is it over-provisioned?"
  • "Analyze the daily memory usage patterns of my staging server to see if it's a good candidate for auto-scaling."

Designed for safety and control

We understand that giving AI agents access to your infrastructure requires trust. The Render MCP server is designed with safety as a core principle. While AI agents are powerful, they aren't guaranteed to make the right decision every time. To protect your services, we have intentionally limited the scope of destructive actions. The server does not support deleting services or databases. The primary sensitive operation it supports is updating a service's environment variables, giving you powerful configuration control without exposing you to unnecessary risk.

Get started in minutes

Connecting your favorite tool to Render is a simple three-step process.

1. Create a Render API key: Generate a new API key from your Account Settings page. Remember to keep this key secure.
2. Configure your tool: Connect your tool to Render's hosted MCP server.
   • We have step-by-step instructions for Claude Code, VS Code, Zed, and more in our official documentation.
   • For example, to configure Cursor, you would add the following to your ~/.cursor/mcp.json file:
   jsonCopy to clipboard

   {
     "mcpServers": {
       "render": {
         "url": "https://mcp.render.com/mcp",
         "headers": {
           "Authorization": "Bearer <YOUR_API_KEY>"
         }
       }
     }
   }

3. Restart your tool: For the changes to take effect, you may need to restart your IDE or command-line tool.

You're ready to go!

Now you're prompting with power

The Render MCP server is a secure and powerful way to integrate your cloud infrastructure directly into your AI development workflows. Explore the full list of capabilities in our documentation and see how you can build, debug, and optimize Render infrastructure faster than ever. For feature requests and contributions, visit the server's public GitHub repository.