BGE
Embedding model family that ranks well on retrieval benchmarks.
Overview
BGE is a family of embedding models trained specifically for retrieval, available in sizes small enough to run locally. It solves the problem that a general-purpose embedding model is not optimised for the asymmetric comparison a search query and a document require. It exists because retrieval quality is set at the embedding step, so a model trained for it improves everything downstream without changing anything else.
Technology Graph
How BGE connects to the rest of the stack.
Tools
Software that implements or integrates with BGE.
Benchmarks
LLM Frameworks
Alternatives
-
Cohere Embed
model
An alternative to BGE · source
-
E5
model
An alternative to BGE · source
-
GTE
model
An alternative to BGE · source
-
Instructor
model
An alternative to BGE · source
-
Jina Embeddings
model
An alternative to BGE · source
-
Nomic Embed
model
An alternative to BGE · source
-
OpenAI Text Embedding 3
model
An alternative to BGE · source
Comparisons
Side-by-side breakdowns with the alternatives above.