WildClawBench, longest of three requests
GPT 6 Astra on “Create a first-half football report with correctly aligned video clips”
Within 5%
Within 5%Worked 103% of the time asked, within 5% of it.
- Asked for
- 30 minutes
- Worked
- 30 min 48 s
- Grade
- Score 17.4 of 100
- Judged afterwards.
- How it ended
- Ended on its own
Every detail of this run
- Agent
- GPT 6 Astra, in Codex
- Benchmark
- WildClawBench
- Asked for
- 30 minutes: “Please work on this task for a full 30 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 30 min 48 s (1,848 s)
- Ratio
- 1.03× the request
- How it ended
- Ended on its own
- Grade
- 0.174
- Task grader score, 0 to 1 (a model judge only where the task calls for one). No official pass rule.
- Score
- 17.4 of 100
- Judged afterwards. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- g3pakxsxk5
- Data release
- 26 Sep 2026
Related runs
| Astra on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest | 2 minutes | 2 min 48 s | 1.40× longer | |
| Middle | 8 minutes | 8 min 17 s | Within 5% | |
| Longest (this run) | 30 minutes | 30 min 48 s | Within 5% | |
| Other agents, longest request | ||||
| GPT 5.6 Sol | 30 minutes | 30 min 28 s | Within 5% | |
| Claude Fable 5.1 | 30 minutes | 22 min 58 s | 1.31× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Ding et al., 2026