Model Library / Benchmarks / MTEB

Retrieval

MTEB

Massive Text Embedding Benchmark — retrieval, clustering, and similarity across dozens of tasks; for embedding models, not chat models.

Features measured

Embeddings

Use cases unlocked

Semantic search RAG retrieval Similarity & clustering

Leaderboard

Higher is better · avg score
1 E5 Mistral 7B (embeddings) 66.6

Benchmark source ↗

← All benchmarks