Unichat MCP Server in Python

Also available in TypeScript

--

Send requests to OpenAI, Anthropic, and OpenAI-compatible providers using MCP protocol via tool or predefined prompts. For OpenAI-compatible providers such as MistralAI, xAI, Google AI, DeepSeek, Alibaba, or Inception, set `UNICHAT_BASE_URL` to the provider's compatible API endpoint.

Vendor API key required

Tools

The server implements one tool:

• `unichat`: Send a request to unichat
  • Takes "messages" as required string arguments
  • Returns a response

Prompts

• `code_review`
  • Review code for best practices, potential issues, and improvements
  • Arguments:
    • `code` (string, required): The code to review"
• `document_code`
  • Generate documentation for code including docstrings and comments
  • Arguments:
    • `code` (string, required): The code to comment"
• `explain_code`
  • Explain how a piece of code works in detail
  • Arguments:
    • `code` (string, required): The code to explain"
• `code_rework`
  • Apply requested changes to the provided code
  • Arguments:
    • `changes` (string, optional): The changes to apply"
    • `code` (string, required): The code to rework"

Quickstart

Install

Claude Desktop

On MacOS: `~/Library/Application\ Support/Claude/claude_desktop_config.json`

On Windows: `%APPDATA%/Claude/claude_desktop_config.json`

Supported Models:

> A list of currently supported models to be used as `"SELECTED_UNICHAT_MODEL"` may be found here. Please make sure to add the relevant vendor API key as `"YOUR_UNICHAT_API_KEY"`

Example:

"env": {
  "UNICHAT_MODEL": "gpt-5.4-mini",
  "UNICHAT_API_KEY": "YOUR_OPENAI_API_KEY"
}

For OpenAI-compatible providers with custom endpoints:

"env": {
  "UNICHAT_MODEL": "PROVIDER_MODEL",
  "UNICHAT_API_KEY": "YOUR_PROVIDER_API_KEY",
  "UNICHAT_BASE_URL": "https://provider.example.com/v1"
}

When `UNICHAT_BASE_URL` is set, the server accepts the configured `UNICHAT_MODEL` without checking it against Unichat's built-in model list.

Development/Unpublished Servers Configuration

"mcpServers": {
  "unichat-mcp-server": {
    "command": "uv",
    "args": [
      "--directory",
      "{{your source code local directory}}/unichat-mcp-server",
      "run",
      "unichat-mcp-server"
    ],
    "env": {
      "UNICHAT_MODEL": "SELECTED_UNICHAT_MODEL",
      "UNICHAT_API_KEY": "YOUR_UNICHAT_API_KEY"
    }
  }
}

Published Servers Configuration

"mcpServers": {
  "unichat-mcp-server": {
    "command": "uvx",
    "args": [
      "unichat-mcp-server"
    ],
    "env": {
      "UNICHAT_MODEL": "SELECTED_UNICHAT_MODEL",
      "UNICHAT_API_KEY": "YOUR_UNICHAT_API_KEY"
    }
  }
}

Installing via Smithery

To install Unichat for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install unichat-mcp-server --client claude

Development

Building and Publishing

To prepare the package for distribution:

1. Remove older builds:

rm -rf dist

2. Sync dependencies and update lockfile:

uv sync

3. Build package distributions:

uv build

This will create source and wheel distributions in the `dist/` directory.

4. Publish to PyPI:

uv publish --token {{YOUR_PYPI_API_TOKEN}}

Debugging

Since MCP servers run over stdio, debugging can be challenging. For the best debugging

experience, we strongly recommend using the MCP Inspector.

You can launch the MCP Inspector via `npm` with this command:

npx @modelcontextprotocol/inspector uv --directory {{your source code local directory}}/unichat-mcp-server run unichat-mcp-server

Upon launching, the Inspector will display a URL that you can access in your browser to begin debugging.

Hosted deployment

A hosted deployment is available on Fronteir AI.