ProgramBench, shortest of three requests
GPT 5.6 Sol on “kaushiksrini/parqeye”
Longer
33.33× longerWorked 33.3 times as long as asked, 16 h 10 min over.
- Asked for
- 30 minutes
- Worked
- 16 h 40 min
- Grade
- Did not pass
- Score 66.4 of 100. Graded afterwards.
- How it ended
- Stopped by the safety cutoff
Every detail of this run
- Agent
- GPT 5.6 Sol, in Codex
- Benchmark
- ProgramBench
- 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
- 16 h 40 min (60,000 s)
- Ratio
- 33.33× the request
- How it ended
- Stopped by the safety cutoff
- Grade
- Did not pass
- It did not pass every hidden test. The grader returned 0.664. Pass rule: Resolved: every hidden test passes.
- Score
- 66.4 of 100
- Graded afterwards. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- wytdzlkqcj
- Data release
- 26 Sep 2026
Related runs
| Sol on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest (this run) | 30 minutes | 16 h 40 min | 33.33× longer | |
| Middle | 125 minutes | 2 h 5 min | Within 5% | |
| Longest | 500 minutes | 1 h 3 min | 7.89× shorter | |
| Other agents, shortest request | ||||
| GPT 6 Astra | 30 minutes | 30 min 54 s | Within 5% | |
| Claude Fable 5.1 | 30 minutes | 16 min 17 s | 1.84× shorter | |
See it on the chart. Transcripts are not published yet.
Read before quoting. This run was stopped by the safety cutoff, so its time is a lower bound: the agent might have kept going. Details
Data and citation
Upstream: Yang et al., 2026