Deployment Guide

Deploy your MCP servers to the ModelFetch platform with built-in hosting, authentication, and scaling. Get your server live in minutes.

Prerequisites: Before deploying, make sure you have a ModelFetch account and have created an organization. You'll also need to create an MCP server instance through the web dashboard.

Deployment Process

Deploying your MCP server requires just two commands:

# First, login to your account (one time only)
modelfetch login

# Second, deploy your MCP server
npm run deploy

The deploy command handles everything automatically - just select your organization and MCP server when prompted.

Environment Variables

Configure environment variables for your deployed MCP server through the ModelFetch web dashboard:

Navigate to your MCP server in the web dashboard
Go to the "Environment Variables" section
Add your production environment variables
Deploy your server to apply the changes

Security: Environment variables are encrypted at rest and in transit. Use them for API keys, database URLs, and other sensitive configuration.

Best Practices

Before deploying:

Test your MCP server locally with production build
Verify all environment variables are configured
Ensure your code handles errors gracefully
Review your dependencies for security vulnerabilities
Test with actual AI clients to verify functionality

Troubleshooting

Common deployment issues and solutions:

"No MCP servers found"

Create an MCP server instance through the web dashboard before deploying.

"Authentication failed"

Run modelfetch logout then modelfetch login to refresh your authentication.

"Build failed"

Ensure your code builds locally with npm run build before deploying.

Next Steps

After successfully deploying your MCP server:

Connect AI clients to your deployed MCP server URL
Monitor performance and usage through the dashboard
Set up environment variables for different environments
Explore advanced CLI features for more deployment options

Congratulations! Your MCP server is now live and ready to power AI applications. The ModelFetch platform handles hosting, scaling, and security so you can focus on building great tools.