All Tools
M
RAGFreeOpen Source
MTEB
The standard benchmark for evaluating text embedding models
Apache-2.0
ABOUT
The embedding model landscape is crowded and rapidly evolving — dozens of new models are released weekly, each claiming state-of-the-art performance on different benchmarks measured by different methodologies. Without a standardized evaluation framework, it is impossible to objectively compare embedding models for retrieval, clustering, classification, or semantic search tasks. MTEB solves this by providing a unified, community-maintained benchmark spanning over 100 datasets and 7 task categories — enabling researchers and practitioners to compare embedding models on equal footing, with reproducible scoring and standardized leaderboards.
INSTALL
pip install mtebINTEGRATION GUIDE
1. Compare and select the best text embedding model for a specific task using standardized evaluation metrics
2. Benchmark new embedding models against community baselines across 100+ diverse datasets
3. Validate embedding quality for RAG pipelines by evaluating retrieval and reranking performance
4. Track embedding model improvements over time with reproducible evaluation pipelines
5. Research embedding model strengths and weaknesses across different languages, domains, and task types
TAGS
pythonevaluationembeddingsbenchmarkretrievalragnlp