reasoning benchmark
TAU-bench Airline
Part of τ-bench (TAU-bench), a benchmark for Tool-Agent-User interaction in real-world domains. The airline domain evaluates language agents' ability to interact with users through dynamic conversations while following domain-specific rules and using API tools. Agents must handle airline-related tasks and policies reliably.
Updated Aug 7, 2026
Published models23
Registry coverage23
MetricScore
EvidenceB
TAU-bench Airline leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | AN | 70.0% | 100.0% | 23 | C | |
| 2 | MI | 62.0% | 95.5% | 23 | C | |
| 3 | ZA | 60.8% | 90.9% | 23 | C | |
| 4 | ZA | 60.4% | 86.4% | 23 | C | |
| 5 | AN | 60.0% | 81.8% | 23 | C | |
| 6 | MI | 60.0% | 77.3% | 23 | C | |
| 7 | AC | 60.0% | 72.7% | 23 | C | |
| 8 | AN | 59.6% | 68.2% | 23 | C | |
| 9 | AN | 58.4% | 63.6% | 23 | C | |
| 10 | AN | 56.0% | 59.1% | 23 | C | |
| 11 | OP | 50.0% | 54.5% | 23 | C | |
| 12 | OP | 50.0% | 50.0% | 23 | C | |
| 13 | OP | 49.4% | 45.5% | 23 | C | |
| 14 | OP | 49.2% | 40.9% | 23 | C | |
| 15 | AC | 49.0% | 36.4% | 23 | C | |
| 16 | AN | 46.0% | 31.8% | 23 | C | |
| 17 | AC | 46.0% | 27.3% | 23 | C | |
| 18 | AC | 44.0% | 22.7% | 23 | C | |
| 19 | OP | 42.8% | 18.2% | 23 | C | |
| 20 | OP | 36.0% | 13.6% | 23 | C | |
| 21 | OP | 32.4% | 9.1% | 23 | C | |
| 22 | AN | 22.8% | 4.5% | 23 | C | |
| 23 | OP | 14.0% | 0.0% | 23 | C |
Score distribution
Top published rows on the benchmark's original scale.
TAU-bench Airline highlights
The top published results on this benchmark's own scale.
What is TAU-bench Airline?
Definition and scoring fields from the benchmark registry.
Part of τ-bench (TAU-bench), a benchmark for Tool-Agent-User interaction in real-world domains. The airline domain evaluates language agents' ability to interact with users through dynamic conversations while following domain-specific rules and using API tools. Agents must handle airline-related tasks and policies reliably.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- TAU-bench Airline
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- tau-bench-airline|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about TAU-bench Airline.
Which model scores highest on TAU-bench Airline?
Claude Sonnet 4.5 is currently ranked first with 70.0%.
What does TAU-bench Airline measure?
Part of τ-bench (TAU-bench), a benchmark for Tool-Agent-User interaction in real-world domains. The airline domain evaluates language agents' ability to interact with users through dynamic conversations while following domain-specific rules and using API tools. Agents must handle airline-related tasks and policies reliably.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
23 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.