What is the current top score on Taubench?

The top reported score is 61.2 by GPT-4o, across 12 models reporting (7 from frontier labs).

What license is Taubench under?

Taubench is available under unknown.

Frontier

tau-bench challenges agents to coordinate, guide, and assist users in achieving shared objectives across complex enterprise domains.

Cite

Notes

Only stored in your browser.

Attribution

Top score 61.2 by GPT-4o - 12 models reporting (7 frontier)

TaubenchBar chart with 12 bars. Highest value: GPT-4o at 61.2.

12 models

What is Taubench?: tau-bench challenges agents to coordinate, guide, and assist users in achieving shared objectives across complex enterprise domains.
What is the current top score on Taubench?: The top reported score is 61.2 by GPT-4o, across 12 models reporting (7 from frontier labs).
What license is Taubench under?: Taubench is available under unknown.