OSWorld-Science Benchmark

What is OSWorld-Science?

Scientific software presents a demanding test for computer-using agents based on visual language models (VLMs): completing a research workflow requires…

Released

2026-09-30

Evaluates

Agents, Agents & Tool Use, General AI

Openness

unknown

Importer

Claire Radar

Review status

rule-auto-admitted

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2609.39903

OSWorld-Science paper and code

• Paper https://arxiv.org/abs/2609.39903
• Code https://github.com/DiscoAILab/OSWorld-Science
• Website https://discoailab.github.io/osworld-science-page/

Which sources cite OSWorld-Science?

• OSWorld-Science

Related Agents, Agents & Tool Use, General AI benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 2d-fet-bench
• 3DCodeBench
• 5-Dialects-BN
• A Real-World CAD Segmentation Benchmark
• A3S-Bench
• AARRI-Bench
• ABACUS