ModelCards
← All benchmarks

TauBench V3 (average)

Customer-service style tasks where the model uses tools and follows policies over a conversation with a simulated user.

Your workloadRAG over docs · 12K in · 600 out · 50K requests/mo · 30% cached
  1. 1NVIDIA Nemotron 3 Ultra · NVIDIAbest value
    70.9%

“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

  • NVIDIA Nemotron 3 Ultra Table 10, p.38

    NVIDIA's own runs of the BF16 model via the NeMo Evaluator SDK. Competitor numbers come from the same table. IOI 2025 is a score out of 600. Blank = not reported.

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