ModelCards
← All benchmarks

BEIR (English)

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. 1Nomic Embed Text v2 · Nomic
    52.86 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

  • Nomic Embed Text v2 report Table 1 (also on model card)

    Nomic's own runs (nDCG@10, 512-token inputs, FlagEmbedding); mE5 numbers taken from its paper. BEIR is the retrieval subset of MTEB. Nomic publishes no full MTEB average for v2.

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