ModelCards
← All benchmarks

OpenLogo (mAP)

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. 1Marengo 3.0 · TwelveLabs
    79.2%

“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

  • Marengo 3.0 launch post

    TwelveLabs' own evaluation of the core embedding model (not the Search API). Vertex is Google Vertex multimodal embeddings (1,408 d); Nova is Amazon Nova (3,072 d). Recall metrics are R@1/5/10 averages unless marked mAP.

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