Retrieval and Retrieval-augmented LLMs
About this project
FlagEmbedding is an open-source retrieval and knowledge project on GitHub: Retrieval and Retrieval-augmented LLMs. It lives in retrieval and knowledge management, the layer that lets a model answer from your own documents instead of from training data alone. It is maintained by FlagOpen and published as a public repository you can read, fork and run yourself.
On GitHub the repository is tagged embeddings, information-retrieval, llm, retrieval-augmented-generation, sentence-embeddings and text-semantic-similarity, which is a fair summary of what it covers. There is normally an ingestion pipeline, an embedding and vector index, and a query layer on top. Most of the engineering effort goes into chunking, ranking and keeping the index fresh as documents change.
It is written mainly in Python and released under MIT, a permissive license — commercial use is fine as long as copyright notices are kept. It is actively maintained: the most recent commit landed on October 27, 2026.
To try it, clone the repository and follow its README; that file is the canonical install guide. There is also a project page at http://www.bge-model.com/ for documentation or a hosted demo. Because it runs on your own infrastructure, budget time for environment setup: dependencies, model weights and configuration usually take longer than the first successful run.
Key facts as recorded on GitHub: 12,343 stars; 951 forks; primary language Python; license MIT; last commit on October 27, 2026; topics embeddings, information-retrieval, llm, retrieval-augmented-generation, sentence-embeddings, text-semantic-similarity.