Novita MCP Server logo

Novita MCP Server

by novitalabs

The `novita-mcp-server` is a Model Context Protocol (MCP) server designed to facilitate interaction with Novita AI platform resources. It enables users to manage their Novita AI resources seamlessly through compatible MCP clients like Claude Desktop and Cursor.

View on GitHub

Last updated: N/A

What is Novita MCP Server?

The Novita MCP Server is a tool that allows users to manage Novita AI resources, specifically GPU instances, through a standardized Model Context Protocol. It acts as a bridge between MCP clients and the Novita AI platform.

How to use Novita MCP Server?

To use the server, install it via Smithery or npm. Configure your MCP client (Claude Desktop or Cursor) with the provided JSON configuration, ensuring your Novita API key is set. Then, use natural language commands within the client to manage your GPU instances.

Key features of Novita MCP Server

  • GPU instance management

  • Cluster/Region listing

  • Product listing

  • Template management

  • Container Registry Auth management

  • Network Storage management

Use cases of Novita MCP Server

  • Managing GPU resources directly from Claude Desktop

  • Automating GPU instance creation and deletion

  • Listing available GPU products and regions

  • Managing container registry authentication

  • Managing network storage for AI workloads

FAQ from Novita MCP Server

What is MCP?

MCP stands for Model Context Protocol. It's a standardized way for applications to interact with AI models and related resources.

Which resources are currently supported?

Currently, only GPU instance management is supported. More resource types will be added in future releases.

How do I get my Novita API key?

You can obtain your Novita API key from the Novita AI Key Management page on the Novita AI platform.

Which MCP clients are compatible?

The server is designed to work with Claude Desktop, Cursor, and any other compatible MCP client.

How do I run tests?

Use the command npm test to run the tests located in the src/tests directory.