OpenRouter Tau2-Bench Airline — leaderboard
OpenRouter's own τ²-Bench Airline run: 50 multi-turn airline customer-service tasks where the model must use tools and follow policy, executed against the production endpoints it serves.
Metric: Accuracy (%). Source: openrouter.ai. Status: years away from saturation. 108 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Fable 5 | 82 |
| 2 | Nova Micro | 80.7 |
| 3 | Gemini 3 Flash (Preview) | 79.3 |
| 4 | Qwen 3.5 397B A17B | 78.4 |
| 5 | GPT-5.5 | 78 |
| 6 | Claude Opus 4.8 | 78 |
| 7 | Qwen 3.5 122B A10B | 78 |
| 8 | GPT-5.6 Sol | 77.3 |
| 9 | Claude Opus 5 | 77.3 |
| 10 | Step 3.7 Flash | 77.3 |
| 11 | Nemotron 3 Ultra | 76.9 |
| 12 | Claude Sonnet 4.6 | 76.7 |
| 13 | Claude Opus 4.6 | 76.7 |
| 14 | Gemma 4 31B | 76.7 |
| 15 | Claude Opus 4.5 | 76.7 |
Interactive version: theaggregate.ai/benchmark?slug=openrouter-tau2-bench-airline · How It Works · Data refreshed daily, snapshot 2026-08-06.