FlagEmbedding
Embedding and reranking library from BAAI — home of the BGE family of state-of-the-art retrieval models with multilingual, multimodal, and long-context variants.
embedding-frameworksRecently released
56
Hero Score
Popularity
64
Performance
30
Ecosystem
50
Maturity
69
Dev Experience
68
⭐ 12,111 stars⬇ 111.9K downloads/wkFirst release: Aug 2023Last release: Aug 2026
Async Support: NoPlugin Extensions: MediumSpeed: MediumDoc Quality: HighLearning Curve: Medium
Pros
- • BGE models consistently top MTEB leaderboards across English, Chinese, and multilingual benchmarks
- • Covers the full retrieval stack — dense, sparse, ColBERT, and cross-encoder rerankers in one library
- • Long-context (BGE-M3) and visual (BGE-VL) variants built in
Cons
- • API surface is busier than sentence-transformers — multiple model classes per use case
- • PyTorch dependency makes it heavy compared to ONNX alternatives
- • Documentation is research-paper-style rather than tutorial-driven