#1313th of 34 in Open models
flagembedding
BGE embedding and reranker models with inference, fine-tuning and evaluation code
Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 12.3k
- License
- MIT
- Last commit
- Aug 2026
- Last release
- Aug 2026
- Language
- Python
11,385 on 31 Mar 202612,261 on 20 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
FlagEmbedding is the Python package and code repository for BAAI's BGE embedding and reranker models. It covers inference, fine-tuning and evaluation for dense, sparse and multi-vector retrieval. The model list includes BGE-M3 (100+ languages, 8192 tokens), bge-en-icl and several cross-encoder rerankers. Tutorials and research code are included.
Who it is for: Engineers building search and RAG who need open embedding and rerank models
Strengths
- Embedders and rerankers with inference, fine-tuning and evaluation scripts in one package
- BGE-M3 supports dense, sparse and multi-vector retrieval, 100+ languages, 8192-token input
- MIT license; models published on Hugging Face and package installable via pip
- Includes hard-negative mining, fine-tune scripts and retrieval/RAG tutorials
Weaknesses
- Library only: no server, UI or vector store, so retrieval pipeline is yours to wire
- README is a long news log; some research code lives in separate repos
- Fine-tuning needs the extra [finetune] install; hardware requirements are unknown
- Query instructions differ per model and must be set correctly
What it needs
- GPU optional
- Models: BGE embedding models (Hugging Face), BGE rerankers, BGE-M3, bge-en-icl, bge-multilingual-gemma2
Also in Open models
See all 34| Rank | Project | Score |
|---|---|---|
| 1 | llmfitRust tool that checks which LLMs fit your RAM, CPU and GPU | 87 out of 100 |
| 2 | layaSingle-pass multilingual classifier for typed choice, score and yes/no decisions | 84 out of 100 |
| 3 | timesfmPretrained time-series foundation model for zero-shot forecasting from Google Research | 75 out of 100 |
| #4 | speechPyTorch toolkit for training and deploying ASR, TTS and speech LLM models | 70 out of 100 |
| #5 | yueLyrics-to-song model that writes an editable melody and chord score first | 65 out of 100 |