Skip to content

OSWorld

A computer-use benchmark for GUI task completion across the broader OSWorld task suite.

2 models

Higher is better.

  1. 1Claude Opus 4.5Anthropic66.3
  2. 2Nemotron 3 Nano Omni 30B A3BNVIDIA47.4
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Interactive GUI evaluation
Tasks
Computer-use tasks
Authors
Z.AI (2026)

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.