ProgramBench, longest of three requests
GPT 5.6 Sol on “kaushiksrini/parqeye”
Shorter
7.89× shorterStopped at 13% of the time asked, 7 h 17 min short.
- Asked for
- 500 minutes
- Worked
- 1 h 3 min
- Grade
- Did not pass
- Score 66.6 of 100. Graded afterwards.
- How it ended
- Stopped by an error
Every detail of this run
- Agent
- GPT 5.6 Sol, in Codex
- Benchmark
- ProgramBench
- Asked for
- 500 minutes: “Please work on this task for a full 500 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 1 h 3 min (3,800 s)
- Ratio
- 0.13× the request
- How it ended
- Stopped by an error
- Grade
- Did not pass
- It did not pass every hidden test. The grader returned 0.666. Pass rule: Resolved: every hidden test passes.
- Score
- 66.6 of 100
- Graded afterwards. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- usvblpmigq
- Data release
- 26 Sep 2026
Related runs
| Sol on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest | 30 minutes | 16 h 40 min | 33.33× longer | |
| Middle | 125 minutes | 2 h 5 min | Within 5% | |
| Longest (this run) | 500 minutes | 1 h 3 min | 7.89× shorter | |
| Other agents, longest request | ||||
| GPT 6 Astra | 500 minutes | 8 h 21 min | Within 5% | |
| Claude Fable 5.1 | 500 minutes | 32 min 36 s | 15.34× shorter | |
See it on the chart. Transcripts are not published yet.
Read before quoting. This run was stopped by an error, so its time is a lower bound: the agent might have kept going. Details
Data and citation
Upstream: Yang et al., 2026