BGE-M3
An open-source multimodal embedding model released by the Beijing Academy of Artificial Intelligence (BAAI) that supports vectorization of images, text, and audio. It is licensed under MIT for free commercial use and ranks among the top embedding models in global benchmarks.
Core Facts
Timeline (last 90 days)
Qwen3-Embedding, EmbeddingGemma 2, and nomic-embed-text v1.5 are all Apache 2.0 licensed, BGE-M3 is MIT licensed, and all can be used directly in commercial products; EmbeddingGemma v1 is governed by the Google Gemma terms; jina-embeddings-v5-text-small uses the CC BY-NC 4.0 license, and commercial use requires obtaining a separate license.
Parameter count does not reliably predict indexing speed: EmbeddingGemma-300M indexes more slowly than the larger BGE-M3 (315.0 vs 433.0 chunks per second).
The project is named mini-Rag, open-sourced on GitHub, with the processing pipeline: BGE-M3 → Dense+Sparse Embeddings → Qdrant → Hybrid Retrieval/RRF → BGE Reranker → Relevance Gate → Context Quality Filter → Context Builder → Local LLM.
The BGE-M3 model can simultaneously generate Dense vectors and Sparse vectors; dense vectors excel at capturing semantic similarity, while sparse vectors preserve precise keyword matching capability.
The Beijing Academy of Artificial Intelligence released the BGE-M3 model in January, and the article considers it to be in the top tier globally.
BGE-M3 follows the MIT open-source license, can be deployed locally, supports vectorization of multiple modalities including images, text, and audio, and is completely free.
The three M's in the name BGE-M3 stand for Multi-Linguality, Multi-Functionality, and Multi-Granularity.
This project uses the BGE-M3 model for text embedding.
BGE-M3 was developed by the Beijing Academy of Artificial Intelligence (BAAI), supports over 100 languages, can simultaneously generate dense vectors, sparse vectors, and ColBERT-style multi-vector representations, and supports a maximum input length of 8192 tokens.
BGE-M3 was developed by the Beijing Academy of Artificial Intelligence (BAAI); the three M's in its name stand for Multi-Linguality, Multi-Functionality, and Multi-Granularity, and it supports more than 100 languages.
2 more timeline events
All Facts (13)
BGE-M3 is a multilingual, multi-granularity, multi-functional Embedding model open-sourced by the Beijing Academy of Artificial Intelligence, supporting more than 100 languages.
75%UnverifiedBGE-M3 was developed by the Beijing Academy of Artificial Intelligence (BAAI); the three M's in its name stand for Multi-Linguality, Multi-Functionality, and Multi-Granularity, and it supports more than 100 languages.
60%UnverifiedQwen3-Embedding, EmbeddingGemma 2, and nomic-embed-text v1.5 are all Apache 2.0 licensed, BGE-M3 is MIT licensed, and all can be used directly in commercial products; EmbeddingGemma v1 is governed by the Google Gemma terms; jina-embeddings-v5-text-small uses the CC BY-NC 4.0 license, and commercial use requires obtaining a separate license.
50%UnverifiedParameter count does not reliably predict indexing speed: EmbeddingGemma-300M indexes more slowly than the larger BGE-M3 (315.0 vs 433.0 chunks per second).
50%UnverifiedThe project is named mini-Rag, open-sourced on GitHub, with the processing pipeline: BGE-M3 → Dense+Sparse Embeddings → Qdrant → Hybrid Retrieval/RRF → BGE Reranker → Relevance Gate → Context Quality Filter → Context Builder → Local LLM.
50%UnverifiedThe BGE-M3 model can simultaneously generate Dense vectors and Sparse vectors; dense vectors excel at capturing semantic similarity, while sparse vectors preserve precise keyword matching capability.
50%UnverifiedThe Beijing Academy of Artificial Intelligence released the BGE-M3 model in January, and the article considers it to be in the top tier globally.
50%UnverifiedBGE-M3 follows the MIT open-source license, can be deployed locally, supports vectorization of multiple modalities including images, text, and audio, and is completely free.
50%UnverifiedThe three M's in the name BGE-M3 stand for Multi-Linguality, Multi-Functionality, and Multi-Granularity.
50%UnverifiedBGE-M3 was developed by the Beijing Academy of Artificial Intelligence (BAAI), supports over 100 languages, can simultaneously generate dense vectors, sparse vectors, and ColBERT-style multi-vector representations, and supports a maximum input length of 8192 tokens.
50%UnverifiedThis project uses the BGE-M3 model for text embedding.
50%UnverifiedBGE-M3 can simultaneously generate dense vectors, sparse vectors, and ColBERT-style multi-vector representations, with a maximum input length of 8,192 tokens.
50%UnverifiedBGE和E5系列在MTEB基准测试上表现出色，其中BGE-M3支持多语言、多粒度和多功能检索
50%