Cross-LLM MCP Server

Example Use Cases

• 1

  Gather multiple perspectives on a single prompt by querying several LLM providers in parallel.

• 2

  Compare outputs, model behavior, or token usage across different LLM vendors to select the best fit.

• 3

  Fallback redundancy or mixed-provider orchestration to improve reliability and manage cost across providers.

Description

This open-source MCP server provides tools to call individual LLM providers (OpenAI ChatGPT, Anthropic Claude, DeepSeek, Google Gemini, and xAI Grok) and a combined tool to query all providers with the same prompt. It returns model info, usage statistics, and includes error handling for missing keys, rate limits, and network errors. The server runs over stdio (works with Claude Desktop and other MCP clients) and requires provider API keys configured via environment variables. Benefits include multi-perspective analysis, model comparison, redundancy, and potential cost optimization by choosing providers per use case.