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. Status: saturation imminent. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3.5 397B A17B | 84.4 |
| 2 | GPT-5.2 | 81.6 |
| 3 | Claude Opus 4.5 | 79.6 |
| 4 | Gemini 3 Flash | 76.8 |
| 5 | Gemini 3 Pro | 75.9 |
| 6 | GLM-5 | 73.7 |
| 7 | Claude Sonnet 4.5 | 72.4 |
Interactive version: theaggregate.ai/benchmark?slug=tau3-bench-retail · How It Works · Data refreshed daily, snapshot 2026-09-05.