sierra-research

tau2-bench

Recently maintained

τ-Bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains

Stars
2,382

Community interest

Forks
622

Community forks

7-day growth
+45

Observed snapshot change

Language
Python

Repository metadata

Star history & growth

Observed snapshots only. Missing observations are not invented.

2,382 ★49 observations
2026-08-102026-11-03
Inspect recent observations
DateStars
2026-10-052,163
2026-10-062,172
2026-10-072,181
2026-10-082,185
2026-10-092,195
2026-10-102,204
2026-10-112,214
2026-10-122,225
2026-10-132,231
2026-10-142,235
2026-10-152,239
2026-10-162,244
2026-10-172,248
2026-10-182,257
2026-10-192,262
2026-10-202,272
2026-10-212,284
2026-10-222,296
2026-10-232,306
2026-10-242,312
2026-10-252,318
2026-10-262,335
2026-10-272,340
2026-10-282,351
2026-10-292,358
2026-10-302,361
2026-10-312,365
2026-11-012,368
2026-11-022,377
2026-11-032,382

Get the source & start here

Cloning downloads source; follow official setup and dependency instructions.

git clone https://github.com/sierra-research/tau2-bench.git
READMEgithub.dev

About this project

What is tau2-bench?

τ-Bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains

How do I get started?

Clone with git clone https://github.com/sierra-research/tau2-bench.git. Cloning is not a complete installation; follow the official README for dependencies, hosts and usage.

Can I use it commercially?

Current license information: MIT. Review the original terms, dependencies and your intended use.

What do maintenance and growth indicate?

Current status: Recently maintained. Activity is not a security audit. Growth uses collected snapshots; missing baselines are not estimated.