Skip to content

Benchmark

τ³-Banking

Scores are percentages. Higher is better.

Models measured
5
Scores recorded
5
Sources
1
IQ weight
0.80

Score distribution

26.6%
Mean
23.5%
Median
11.7%
Std dev
9.3%
10th percentile
44.3%
90th percentile
9.3%
Min
9.3%44.3%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.80 for chat models, measured over 5 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Ling 3.0 Flash SanteLing-3.0-flash-Sante official evaluation image · highest reasoning tier · temperature 1.0 / top_p 0.95 unless otherwise stated · exact API snapshot and run count not published · official Tau evaluation set · terminal_use tool only· Reported by Ant inclusionAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.44.3%
2Kimi K3max effort · temp 1.0 · top_p 1.0· Reported by Moonshot AI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.33.4%
3Muse Glimmer 30Bhigh reasoning· Reported by Meta· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.23.5%
4Nemotron 3 UltraNeMo Evaluator SDK · TauBench v3· Reported by NVIDIA· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.22.6%
5Nemotron 3.5 Lightning 30B A3Bofficial release table· Reported by NVIDIA· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.9.3%

Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links.