mcp-arena

wjlgatech/mcp-arena

★ 0 stars Python AI/LLM Updated today

Evidence-gated benchmark for LLM tool-use over real MCP servers — the answer must be earned through the tools

View on GitHub → 🔍 Audit Wallet Slippage →

Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "mcp-arena": {
      "command": "uvx",
      "args": [
        "mcp-arena"
      ]
    }
  }
}

Or install with pip: pip install mcp-arena

README Excerpt

**An evidence-gated benchmark for LLM tool-use over REAL MCP servers.** Existing tool-use benchmarks score mocked function signatures. MCP-Arena scores the thing that matters in production: does the model **earn its answer through the tools** — right tool found, right tool chosen among overlapping ones, multi-step chains composed, errors

mcp-arena

Quick Install

README Excerpt

Related AI/LLM Servers