context.vn

tau3 Bench

8 models evaluated

#ModelProviderTypeScore
1Mistral Medium 3.5 128BMistralOpenτ³-bench Telecom
4Qwen3.6 PlusAlibaba · Closedτ³-bench published setup70.7%
5GLM-5.1Z.AI · Open weightτ³-bench published setup70.6%
6Claude Opus 4.5Anthropic · Closedτ³-bench published setup70.2%
7Qwen3.5 397BAlibaba · Open weightτ³-bench published setup68.4%
8Qwen3.6-35B-A3BAlibaba · Open weightτ³-bench published setup67.2%
9Kimi K2.5Moonshot AI · Open weightτ³-bench published setup65.7%
10GLM-5Z.AI · Open weightτ³-bench published setup65.6%