← τ-bench leaderboards

Telecom · Standard · pass^1

τ-bench · 8 models

RankModelScore
1Qwen3.5-397B-A17B97.81
2Claude Opus 4.592.32
3Gemini 3 Flash91.23
4Gemini 3 Pro91.01
5GPT-5.289.69
6GLM-586.84
7Claude Sonnet 4.584.87
8GPT-5.257.2368