BGE Large EN v1.5
The classic English embedding — 10M downloads of MIT-licensed retrieval reliability.
About
BGE-Large-EN-v1.5 (2023, BAAI) is the classic English embedding model: 10M Hugging Face downloads, MIT license, 512-token context, and the checkpoint that defined "good open embeddings" for a generation of RAG systems.
MIT licensed — commercial use fully allowed. A 2023 model: superseded by BGE-M3 on multilingual and by newer models on quality, but the reliability story is unmatched — a decade of tutorials, integrations, and production deployments reference this exact checkpoint. For English RAG where you want the most battle-tested checkpoint, this remains defensible. The 1.5 fixed the similarity-distribution issues of v1.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-claimed at release (2023): leading MTEB — treat as provider-claimed and dated. Independent: 10M downloads of community validation.
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (verified tier). 10M downloads of MIT-licensed reliability — the most battle-tested English embedding checkpoint in existence. Docked two points: (1) 2023-era, superseded on quality by BGE-M3 and newer models; (2) 512-token context. The safe default for English RAG.
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 9.4/10
- Maintenance · recency, releases, issue load
- 4.0/10
- Openness · license + docs
- 9.4/10
- Manual review · sanity read + red-flag check
- 8.0/10
- · HF downloads 10,032,899 (peer-normalized)
- · released 1138 days ago
- · manual review: most battle-tested English embedding checkpoint; 2023-era, superseded on quality; license: MIT