Llm Router

Name: Llm Router
Author: ypollak2

@ypollak2

About Llm Router

Universal LLM router for AI coding tools. Works with Claude Code, Cursor, Codex, Gemini CLI, Copilot and more. Free-first fallback chain keeps costs 70–85% lower.

#mcp #gemini #openai #claude #cost-optimization #llm #anthropic #ollama #litellm #llm-router #mcp-server #claude-code #ai-routing #model-router

Basic information

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "llm-router": {
      "command": "uv",
      "args": [
        "sync",
        "--extra",
        "dev"
      ]
    }
  }
}

Tools

No tools detected

We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.

Overview

What is Llm Router?

Llm Router is a local-first model router that intercepts prompts from AI coding tools like Claude Code, Codex, and Gemini CLI and routes them to the cheapest capable model first, falling back automatically. It helps developers save 35–80% on routine prompts, protect premium quota, and avoid workflow interruptions when providers fail.

How to use Llm Router?

Install with pip install llm-routing, then run llm-router install to set up hooks for your coding tool. Optionally configure API keys via environment variables for external providers and verify connectivity with llm-router health. Once set up, the router works automatically in the background; you can also use MCP tools like llm_query for manual routing.