Hyper-τ-bench makes agent construction the task: a developer agent recovers requirements from a business’s records and client, builds a customer-service agent within a model and cost budget, and is scored on held-out τ-bench-style traffic.
τ³-bench extends the benchmark family with the τ-knowledge and τ-voice tracks, advancing agent evaluation to unstructured knowledge and real-time voice.
A benchmark for evaluating AI agents in knowledge-intensive customer support: a realistic fintech knowledge base of 698 documents paired with tasks requiring multi-step reasoning, policy application, and tool use.
We audited and fixed 50+ tasks across the airline and retail domains, addressing incorrect expected actions, ambiguous instructions, impossible constraints, and missing fallback behaviors.