Skip to content
Fossfits

benchmark.

Listed 4Updated Sep 11, 2026

  1. Rank 1. OSWorld[NeurIPS 2024] OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer EnvironmentsGitHub stars: 3.1k
  2. Rank 2. OpenSTLOpenSTL: A Comprehensive Benchmark of Spatio-Temporal Predictive LearningGitHub stars: 1.1k
  3. Rank 3. OSWorld-V2OSWorld 2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World TasksGitHub stars: 303
  4. Rank 4. XLingEvalCode and Resources for the paper, "Better to Ask in English: Cross-Lingual Evaluation of Large Language Models for Healthcare Queries"GitHub stars: 202