LLM Stats (TAU3-Bench): leaderboard
LLM Stats aggregate of Tau3-Bench agentic customer-service tasks across retail, telecom, airline, and banking-knowledge domains.
Metric: Score (%). Source: llm-stats.com. Saturation forecast: Around January 2027. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | MiMo-V2.5-Pro | 72.9 | #158 |
| 2 | Qwen 3.6 Plus | 70.7 | #120 |
| 3 | GLM-5.1 | 70.6 | #97 |
| 4 | Qwen 3.6 35B A3B | 67.2 | #226 |
| 5 | Granite 4.2 30B | 62 | #428 |
| 6 | Granite 4.2 8B | 58.06 | #635 |
| 7 | Granite 4.2 3B | 45.78 | #780 |
| 8 | Nemotron 3 Ultra | 22.6 | #235 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=llm-stats-tau3-bench · How It Works · Data refreshed daily, snapshot 2026-10-20.