all the models — AI benchmark observatory
← Benchmarks

TAU3-Bench

TAU3-Bench is a benchmark for evaluating general-purpose agent capabilities, testing models on multi-turn interactions with simulated user models, retrieval, and complex decision-making scenarios.

id tau3-bench · max 1 · 9 models reported

#ModelScore
1MiMo-V2.5-Pro
Xiaomi · open
0.73
2Qwen3.6 Plus
Alibaba Cloud / Qwen Team
0.71
3GLM-5.1
Zhipu AI · open
0.71