reasoning benchmark
Tau2 Retail
τ²-bench retail domain evaluates conversational AI agents in customer service scenarios within a dual-control environment where both agent and user can interact with tools. Tests tool-agent-user interaction, rule adherence, and task consistency in retail customer support contexts.
Updated Aug 7, 2026
Published models26
Registry coverage26
MetricScore
EvidenceB
Tau2 Retail leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
Score distribution
Top published rows on the benchmark's original scale.
Tau2 Retail highlights
The top published results on this benchmark's own scale.
What is Tau2 Retail?
Definition and scoring fields from the benchmark registry.
τ²-bench retail domain evaluates conversational AI agents in customer service scenarios within a dual-control environment where both agent and user can interact with tools. Tests tool-agent-user interaction, rule adherence, and task consistency in retail customer support contexts.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- Tau2 Retail
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- tau2-retail|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about Tau2 Retail.
Which model scores highest on Tau2 Retail?
Claude Opus 4.6 is currently ranked first with 91.9%.
What does Tau2 Retail measure?
τ²-bench retail domain evaluates conversational AI agents in customer service scenarios within a dual-control environment where both agent and user can interact with tools. Tests tool-agent-user interaction, rule adherence, and task consistency in retail customer support contexts.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
26 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.