SYS NOMINALLISTINGS 1305 ▲SERVERS 948CLIENTS 103AI AGENTS 240SKILLS 05PLUGINS 03RULES 03EVALS 03
All tools
τ-
EVAL

τ-bench

Benchmark for tool-agent-user interaction: an agent with API tools and policies serves a simulated user in airline and retail domains.

AI & language modelsevalbenchmarktool-useconversational-agents
Visit project
Claim this listing

About τ-bench

Code and Data for Tau-Bench. τ-bench emulates dynamic conversations between a user (simulated by language models) and a language agent provided with domain-specific API tools and policy guidelines, evaluating tool-agent-user interaction across real-world domains such as airline and retail. The repository notes that its tasks are outdated and points to the newer τ³-bench repository for the latest versions. Papers: τ-bench (arXiv:2406.12045) and τ²-Bench (arXiv:2506.07982). Details were read from the publisher’s repository and documentation; nothing was installed or executed.

What you can do

  • Simulated users
  • Domain API tools and policies
  • Airline and retail tasks

Benchmark

Format
Evaluation set or harness
Works with
Not specified by the publisher
License
MIT

How to use it

How do I run τ-bench?

Follow the publisher’s documentation to get the data and run the harness against your agent or model.

What does τ-bench measure?

Simulated users; Domain API tools and policies; Airline and retail tasks.

Is it free?

Open source; free to run yourself.

Questions and answers

What is τ-bench?

τ-bench is an eval in the AI & language models category. Benchmark for tool-agent-user interaction: an agent with API tools and policies serves a simulated user in airline and retail domains.

How do I run τ-bench?

Follow the publisher's documentation at https://arxiv.org/abs/2406.12045 to get the data and run the harness against your agent or model.

Is τ-bench free?

Open source; free to run yourself.

What can τ-bench do?

According to the published details: Simulated users; Domain API tools and policies; Airline and retail tasks.

Where is τ-bench published?

Its homepage is https://github.com/sierra-research/tau-bench and its source repository is https://github.com/sierra-research/tau-bench. RUAGENTIC read these details from Publisher repository on October 1, 2026 and did not install or run the project.

Sources and checks

View the source

Source information collected October 1, 2026.

Badge and embed

Show this listing on your website or README. The badge and the card link back to this page.

Listed on RUAGENTICτ-bench on RUAGENTIC
Markdown badge
[![Listed on RUAGENTIC](https://ruagentic.com/badge/tau-bench.svg)](https://ruagentic.com/tools/tau-bench)
HTML badge
<a href="https://ruagentic.com/tools/tau-bench"><img src="https://ruagentic.com/badge/tau-bench.svg" alt="Listed on RUAGENTIC" height="28"></a>
Markdown card
[![τ-bench on RUAGENTIC](https://ruagentic.com/embed/tau-bench.svg)](https://ruagentic.com/tools/tau-bench)
HTML embed
<iframe src="https://ruagentic.com/embed/tau-bench" width="480" height="150" style="border:0;max-width:100%" loading="lazy" title="τ-bench on RUAGENTIC"></iframe>