txtai

txtai is an embeddings database for semantic search, retrieval-augmented generation, and LLM workflows. Developers use it to build search and RAG apps in Python with minimal setup.

txtai is an all-in-one open-source AI framework for Python, centered on an embeddings database for semantic search. That database combines dense and sparse vector indexes, graph networks, and a relational store, so you can run vector search with SQL, topic modeling, and graph analysis, and index text, documents, audio, images, and video. On top of that foundation it adds pipelines (for tasks like question-answering, summarization, transcription, and translation), workflows to chain those pipelines, and agents that connect everything to autonomously tackle more complex problems. It is built on Hugging Face Transformers, Sentence Transformers, and FastAPI, and is designed to run locally with sensible defaults so you can start in a couple of lines of code. It also exposes web and Model Context Protocol (MCP) APIs with bindings for JavaScript, Java, Rust, and Go. It is popular with Python developers building semantic search, RAG, and LLM-workflow applications without wiring together many separate tools.