LLM.kiwi LogoLLM.kiwi

MCP Server

Integrate LLM.kiwi with AI-powered IDEs using the Model Context Protocol (MCP).

What is MCP?

The Model Context Protocol (MCP) is a standard for connecting AI assistants to external tools and data sources. LLM.kiwi provides an MCP server that enables IDEs like Cursor, Windsurf, and VS Code to use our API directly.

MCP Endpoint

https://api.llm.kiwi/mcp

Available Tools

chat_completion

Generate text using LLM.kiwi chat models

Parameters:

messages: arraymodel?: stringtemperature?: number
image_generation

Generate images from text descriptions

Parameters:

prompt: stringsize?: string
audio_transcription

Transcribe audio files to text

Parameters:

file_url: stringlanguage?: string

IDE Configuration

Cursor

Add to your .cursor/mcp.json:

{
  "mcpServers": {
    "llm-kiwi": {
      "url": "https://api.llm.kiwi/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

VS Code (with MCP Extensions)

Add to your VS Code settings:

{
  "mcpServers": {
    "llm-kiwi": {
      "url": "https://api.llm.kiwi/mcp",
      "apiKey": "YOUR_API_KEY"
    }
  }
}

Resources

The MCP server also exposes resources for AI agents to read:

  • • models://list - Available models and capabilities
  • • docs://api - Full API documentation as markdown
  • • usage://current - Current usage statistics

Example Usage

Once configured, you can use natural language in your IDE:

# In Cursor or MCP-enabled IDE chat:

"Use the llm-kiwi chat_completion tool to write a haiku about coding"