$\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
Explore score history ↗A benchmark where a developer agent builds a complete customer-service agent from business records, client requirements, a production API, inherited codebase, and cost/model limits, scored by deployment against held-out simulated users across 53 tasks in four domains.
- Released
- 2026-09-04
- Institutions
- Princeton University
- Evaluates
- Agents, Agents & Tool Use, General AI
- Openness
- unknown
- Importer
- Claire Radar
- Review status
- ai-reviewed
$\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction paper
No reported scores are on record for this benchmark yet.