Atlas

Benchmarks

← All benchmarks

BankerToolBench v1.1 Pass@3

Professional Work · 2026-10-09

Share of BankerToolBench v1.1's 100 tasks on which at least one of three runs passes every critical-weight rubric criterion. It is a component of the v1.1 factor group.

Top models (higher is better)

ModelScore
Claude Opus 542.0
GPT-5.6 Sol36.0
Opus 4.830.0
GPT-5.524.0
GLM-5.223.0
GPT-5.422.0
Gemini 3.1 Pro Preview20.0
Opus 4.612.0
GLM-511.0
Qwen3.5 397B A17B5.0
Gemini 2.5 Pro2.0
Loading Atlas data…