TAU3-Bench Benchmark Leaderboard | Phaseo

TAU3-Bench

AI benchmark results and model performance

Type: percentageAgents

Recorded Results

6

Average Score

39.53%

Score Range

9.28% - 91.40%

Leading Model (lowest score)

9.28% - Nemotron 3.5 Lightning 30B A3B

Scores Over Time

Individual benchmark scores plotted by date.

Models Using This Benchmark

Organisation	Model	Reported	Top Score	Info	Self Reported	Source
Nvidia	Nemotron 3.5 Lightning 30B A3B	11 Aug 2026	9.28%	Banking; NVIDIA release evaluation suite	Yes	Source
Sakana AI	Fugu Ultra	22 Jun 2026	20.60%	Sakana reports this benchmark as Tau3 Banking.	Yes	Source
Sakana AI	Fugu	22 Jun 2026	21.70%	Sakana reports this benchmark as Tau3 Banking.	Yes	Source
Meta	Muse Glimmer 30B	10 Aug 2026	23.50%	Banking domain; high reasoning; result sourced by Meta from Artificial Analysis	Yes	Source
Qwen	Qwen 3.6 Plus	01 Apr 2026	70.70%	-	Yes	Source
Mistral	Mistral Medium 3.5	29 Apr 2026	91.40%	Telecom	Yes	Source