A fun experiment to see how well API platforms handle some real-world AI tasks using their services.<p>Along with this we released a new tool use eval framework to help determine accuracy, speed, and model quality.<p><a href="https://docs.mcp.run/blog/2025/03/03/introducing-mcpx-eval" rel="nofollow">https://docs.mcp.run/blog/2025/03/03/introducing-mcpx-eval</a>