Ragtime is a self-hosted RAG API and Model Context Protocol (MCP) server that indexes local knowledge (uploads, Git repos, schemas) and serves it to LLM clients via an OpenAI-compatible `/v1/chat/completions` endpoint with streaming. It supports a wide range of model providers (OpenAI, Anthropic, OpenRouter, Ollama, llama.cpp, LM Studio, and more) with a built-in chat UI and dual vector store options (FAISS or pgvector).
Use this when you need to add private document or codebase knowledge to LLM workflows while keeping data self-hosted, or when you want a unified API layer that works across multiple AI providers and subscription accounts. It's especially valuable for teams building chatbots, internal knowledge tools, or agentic applications that require RAG without depending on a single SaaS vendor.