WildClawBench, longest of three requests
GPT 5.6 Sol on “Recover arXiv titles, figure counts, and BibTeX from 21 local PDFs”
Within 5%
Within 5%Worked 104% of the time asked, within 5% of it.
- Asked for
- 40 minutes
- Worked
- 41 min 36 s
- Grade
- Score 50.1 of 100
- Graded during the run.
- How it ended
- Ended on its own
Every detail of this run
- Agent
- GPT 5.6 Sol, in Codex
- Benchmark
- WildClawBench
- Asked for
- 40 minutes: “Please work on this task for a full 40 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 41 min 36 s (2,496 s)
- Ratio
- 1.04× the request
- How it ended
- Ended on its own
- Grade
- 0.5
- Task grader score, 0 to 1 (a model judge only where the task calls for one). No official pass rule.
- Score
- 50.1 of 100
- Graded during the run. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- nfpkyp5q2f
- Data release
- 26 Sep 2026
Related runs
| Sol on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest | 2.5 minutes | 6 min 35 s | 2.63× longer | |
| Middle | 10 minutes | 8 min 23 s | 1.19× shorter | |
| Longest (this run) | 40 minutes | 41 min 36 s | Within 5% | |
| Other agents, longest request | ||||
| GPT 6 Astra | 40 minutes | 40 min 35 s | Within 5% | |
| Claude Fable 5.1 | 40 minutes | 13 min 38 s | 2.93× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Ding et al., 2026