OSWorld 2.0, shortest of three requests
GPT 6 Astra on “Zotero, Obsidian, and BibTeX”
Longer
1.35× longerWorked 35% past the request, 5 min 11 s over.
- Asked for
- 15 minutes
- Worked
- 20 min 11 s
- Grade
- Did not pass
- Score 99 of 100. Graded during the run.
- How it ended
- Ended on its own
Every detail of this run
- Agent
- GPT 6 Astra, in Codex
- Benchmark
- OSWorld 2.0
- Asked for
- 15 minutes: “Please work on this task for a full 15 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 20 min 11 s (1,211 s)
- Ratio
- 1.35× the request
- How it ended
- Ended on its own
- Grade
- Did not pass
- It did not meet every checkpoint. The grader returned 0.99. Pass rule: Binary completion: a score of 1.00, with every checkpoint met.
- Score
- 99 of 100
- Graded during the run. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- dx42n5mmrq
- Data release
- 26 Sep 2026
Related runs
| Astra on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest (this run) | 15 minutes | 20 min 11 s | 1.35× longer | |
| Middle | 60 minutes | 1 h 14 min | 1.24× longer | |
| Longest | 250 minutes | 4 h 11 min | Within 5% | |
| Other agents, shortest request | ||||
| GPT 5.6 Sol | 15 minutes | 1 h 27 min | 5.79× longer | |
| Claude Fable 5.1 | 15 minutes | 12 min 1 s | 1.25× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Yuan et al., 2026