Discover and compare models and providers with up-to-date pricing, limits, modalities, and capabilities. Search and filter by features like reasoning, tool calling, and context length, then fetch deta
"primitive" RAG-like web search model context protocol (MCP) server that runs locally. No APIs needed.