ModelCards
← All benchmarks

MTEB overall average

We haven't written a description for this benchmark yet; see the card notes below for how each maker ran it.

Your workloadRAG over docs · 12K in · 600 out · 50K requests/mo · 30% cached
  1. 1Amazon Titan Text Embeddings V2 · Amazon
    0.6 score

“Best value”: no other model here reports a higher score at a lower cost for your workload. Scores are self-reported by each maker with different settings, so treat close calls as ties.

How each maker reported it

  • Amazon Titan Text Embeddings V2 service card (Robustness)

    AWS's own numbers on MTEB (0–1 scale as published). Retrieval = average NDCG@10 on MTEB retrieval sets. V2 binary embeddings keep 98.5% of full-precision retrieval (with full-precision query reranking); 512 and 256 dims keep 99.0% and 96.8% of 1024-dim retrieval.

Only each model's own maker's number is shown; scores a maker reports for rival models are left out.