Tau3-Bench Retail: leaderboard
τ³-bench retail domain: multimodal, knowledge-aware agentic customer service tasks. Upgraded from τ²-bench with improved task quality. From Sierra AI.
Metric: Pass@1 (%). Source: taubench.com. Saturation forecast: Not forecast. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.5 397B A17B | 84.4 | #146 |
| 2 | GPT-5.2 | 81.6 | #110 |
| 3 | Claude Opus 4.5 | 79.6 | #83 |
| 4 | Gemini 3 Flash | 76.8 | #99 |
| 5 | Gemini 3 Pro | 75.9 | #80 |
| 6 | GLM-5 | 73.7 | #141 |
| 7 | Claude Sonnet 4.5 | 72.4 | #142 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=tau3-bench-retail · How It Works · Data refreshed daily, snapshot 2026-10-20.