ModelCards
← All benchmarks

WorldVQA (with tools)

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. 1Muse Spark Open · Metabest value
    49.4%

“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

  • Muse Spark Open model card (Benchmarks); evaluation methodology report

    Muse Spark Open at xhigh effort, temperature 1.0, top_p 1.0. Meta compares it only with other open-weight models; settings for the others are as labelled. Coding rows use each model's preferred harness except Terminal-Bench (Terminus 2 agent). Null = not reported.

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