WildClawBench, middle of three requests
GPT 5.6 Sol on “Optimize 30 calendar events plus 15 meeting requests under constraints”
Longer
1.79× longerWorked 79% past the request, 3 min 9 s over.
- Asked for
- 4 minutes
- Worked
- 7 min 9 s
- Grade
- Score 100 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
- 4 minutes: “Please work on this task for a full 4 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 7 min 9 s (429 s)
- Ratio
- 1.79× the request
- How it ended
- Ended on its own
- Grade
- 1
- Task grader score, 0 to 1 (a model judge only where the task calls for one). No official pass rule.
- Score
- 100 of 100
- Graded during the run. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- k7qudcqfsm
- Data release
- 26 Sep 2026
Related runs
| Sol on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest | 1 minute | 7 min 45 s | 7.75× longer | |
| Middle (this run) | 4 minutes | 7 min 9 s | 1.79× longer | |
| Longest | 15 minutes | 11 min 9 s | 1.34× shorter | |
| Other agents, middle request | ||||
| GPT 6 Astra | 4 minutes | 4 min 14 s | 1.06× longer | |
| Claude Fable 5.1 | 4 minutes | 7 min 12 s | 1.80× longer | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Ding et al., 2026