BAAI Model Family
What is BAAI Model Family?
BGE (BAAI General Embedding) is an open-source model family for text embedding and reranking developed by the Beijing Academy of Artificial Intelligence (BAAI). Published as open weights, the models run on self-hosted or customer-managed inference infrastructure using frameworks such as FlagEmbedding and sentence-transformers to convert text into vector representations for semantic search, retrieval-augmented generation (RAG), text clustering, and classification.
The portfolio spans English and multilingual embedding models across small, base, and large sizes, alongside companion cross-encoder rerankers used in two-stage retrieval pipelines. Specialized variants include BGE-M3, which processes inputs up to 8,192 tokens and emits dense, sparse, and multi-vector representations from a single architecture.