Model Radar is an open-source MCP server for AI and coding workflow developers. It pings, catalogs, and benchmarks coding-focused LLMs from various free/public providers and subscription CLIs (like Claude, Grok, Antigravity/Gemini, Codex), ranks them by latency and quality tier, and exposes a local web dashboard and API for fast model selection and batch evaluations. Designed for integration with AI agents and client tools, it enables automated discovery, performance-based selection, and parallel job execution across numerous model backends, while keeping API keys local for privacy.
Visit Model Radar's official website for product details and getting started.
Report bugs or request features directly on the GitHub repository.