Name: OSWorld 2.0 v2026.08.08 Offline — Partial
Creator: Benchmark Registry

Cite

Fill in the access date before citing.

Plain text

Benchmark Registry. OSWorld 2.0 v2026.08.08 Offline — Partial benchmark results. Benchmark identifier: osworld/2.0-v2026.08.08-offline-partial. https://benchmarkregistry.org/benchmarks/osworld/2.0-v2026.08.08-offline-partial. Accessed [YYYY-MM-DD]. Data license: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Attribution: Benchmark Registry, a project of Densa Labs — https://benchmarkregistry.org/ (CC BY 4.0). Include the cited record URL and identify any changes.

BibTeX

@misc{br_osworld_2_0_v2026_08_08_offline_partial,
  title = {OSWorld 2.0 v2026.08.08 Offline — Partial benchmark results},
  publisher = {Benchmark Registry},
  url = {https://benchmarkregistry.org/benchmarks/osworld/2.0-v2026.08.08-offline-partial},
  urldate = {YYYY-MM-DD},
  note = {Benchmark identifier: osworld/2.0-v2026.08.08-offline-partial; data license: CC BY 4.0; access date to be filled by the reader}
}

Report an issue

Benchmark: OSWorld
Evaluated by: XLang Lab
Release date: August 8, 2026
Version: OSWorld 2.0 v2026.08.08 Offline — Partial
Metric: OSWorld Partial Score

Showing the selected evaluation. Show all results

Score

GPT-6 Astra (high)

70.0%

openai.comOSWorld Partial ScoreReported September 22, 2026OpenAIReport an issue

Cite

Fill in the access date before citing.

Plain text

Benchmark Registry. GPT-6 Astra: OSWorld 2.0 v2026.08.08 Offline — Partial, 70.0% (high). Registry No. 10005; Record No. c605ae83f7632c686c70ef0055b8e0ac60cc76a27cf5cdddfdef294dca4a0c3e. https://benchmarkregistry.org/benchmarks/osworld/2.0-v2026.08.08-offline-partial?view=history&result=c605ae83f7632c686c70ef0055b8e0ac60cc76a27cf5cdddfdef294dca4a0c3e#BR-c605ae83f7632c686c70ef0055b8e0ac60cc76a27cf5cdddfdef294dca4a0c3e. Accessed [YYYY-MM-DD]. Data license: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Attribution: Benchmark Registry, a project of Densa Labs — https://benchmarkregistry.org/ (CC BY 4.0). Include the cited record URL and identify any changes.

BibTeX

@misc{br_c605ae83f7632c686c70ef0055b8e0ac60cc76a27cf5cdddfdef294dca4a0c3e,
  title = {GPT-6 Astra: OSWorld 2.0 v2026.08.08 Offline — Partial, 70.0\% (high)},
  publisher = {Benchmark Registry},
  url = {https://benchmarkregistry.org/benchmarks/osworld/2.0-v2026.08.08-offline-partial?view=history\&result=c605ae83f7632c686c70ef0055b8e0ac60cc76a27cf5cdddfdef294dca4a0c3e\#BR-c605ae83f7632c686c70ef0055b8e0ac60cc76a27cf5cdddfdef294dca4a0c3e},
  urldate = {YYYY-MM-DD},
  note = {Registry No. 10005; Record No. c605ae83f7632c686c70ef0055b8e0ac60cc76a27cf5cdddfdef294dca4a0c3e; data license: CC BY 4.0; access date to be filled by the reader}
}

Permalink to this record

Badge for this model's latest reported score in the benchmark family; it may differ from this record.

Markdown

![GPT-6 Astra: OSWorld](https://benchmarkregistry.org/badge/10005/osworld.svg)

HTML

<img src="https://benchmarkregistry.org/badge/10005/osworld.svg" alt="GPT-6 Astra: OSWorld" height="24">

10005

Results