openai/chatgpt-retrieval-plugin

Implements semantic search via embeddings across 13+ vector database backends (Pinecone, Weaviate, Qdrant, Milvus, Redis, PostgreSQL, etc.), exposed as a FastAPI server with configurable document chunking, metadata extraction, and PII detection. Integrates with ChatGPT custom GPTs, function calling, and Assistants API, offering granular control over retrieval parameters like embedding model selection and chunk size that native file uploads don't provide.

21,225 stars. No commits in the last 6 months.