OSWorld 2.0, longest of three requests
Claude Fable 5.1 on “Local event booking QA”
Shorter
2.94× shorterStopped at 34% of the time asked, 2 h 45 min short.
- Asked for
- 250 minutes
- Worked
- 1 h 25 min
- Grade
- Did not pass
- Score 78.8 of 100. Graded during the run.
- How it ended
- Ended on its own
Every detail of this run
- Agent
- Claude Fable 5.1, in Claude Code
- Benchmark
- OSWorld 2.0
- Asked for
- 250 minutes: “Please work on this task for a full 250 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 1 h 25 min (5,099 s)
- Ratio
- 0.34× the request
- How it ended
- Ended on its own
- Grade
- Did not pass
- It did not meet every checkpoint. The grader returned 0.788. Pass rule: Binary completion: a score of 1.00, with every checkpoint met.
- Score
- 78.8 of 100
- Graded during the run. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- geqavefr2u
- Data release
- 26 Sep 2026
Related runs
| Fable on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest | 15 minutes | 12 min 24 s | 1.21× shorter | |
| Middle | 60 minutes | 51 min 54 s | 1.16× shorter | |
| Longest (this run) | 250 minutes | 1 h 25 min | 2.94× shorter | |
| Other agents, longest request | ||||
| GPT 6 Astra | 250 minutes | 4 h 13 min | Within 5% | |
| GPT 5.6 Sol | 250 minutes | 2 h 47 min | 1.50× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Yuan et al., 2026