Custom MCP Server Integration

The v0 API now supports connecting to any custom MCP server, giving teams the ability to integrate external tools and services directly into their v0 workflows. This feature enables programmatic configuration and management of MCP servers alongside the standard v0 API.

How It Works

Teams can register new MCP servers by providing an endpoint URL and authentication details through the v0.mcpServers.create() method:

const server = await v0.mcpServers.create({
  name: "Vercel",
  url: "https://mcp.vercel.com",
})

Once a custom server is configured, it becomes immediately available for use during v0 chat sessions by referencing the server ID:

await v0.chats.create({
  message: "Build a heatmap of my team's Vercel deployments.",
  mcpServerIds: [server.id],
})

Use Cases

This capability opens up new possibilities for enterprises, including:

• Custom integrations: Connect proprietary databases, internal APIs, or specialized services to v0
• Enhanced context: Make domain-specific tools available directly within the AI's reasoning process
• Streamlined workflows: Reduce manual steps by letting v0 access the exact tools and data it needs

Getting Started

Developers can explore the full MCP server API in the v0 API documentation.