ProgramBench, middle of three requests
Claude Fable 5.1 on “facebook/zstd”
Shorter
1.97× shorterStopped at 51% of the time asked, 19 min 43 s short.
- Asked for
- 40 minutes
- Worked
- 20 min 18 s
- Grade
- Did not pass
- Score 61 of 100. Graded afterwards.
- How it ended
- Ended on its own
Every detail of this run
- Agent
- Claude Fable 5.1, in Claude Code
- Benchmark
- ProgramBench
- Task
- facebook/zstd
- 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
- 20 min 18 s (1,218 s)
- Ratio
- 0.51× the request
- How it ended
- Ended on its own
- Grade
- Did not pass
- It did not pass every hidden test. The grader returned 0.61. Pass rule: Resolved: every hidden test passes.
- Score
- 61 of 100
- Graded afterwards. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- c3tsbdpj7n
- Data release
- 26 Sep 2026
Related runs
| Fable on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest | 10 minutes | 8 min 50 s | 1.13× shorter | |
| Middle (this run) | 40 minutes | 20 min 18 s | 1.97× shorter | |
| Longest | 150 minutes | 1 h 1 min | 2.47× shorter | |
| Other agents, middle request | ||||
| GPT 6 Astra | 40 minutes | 40 min 35 s | Within 5% | |
| GPT 5.6 Sol | 40 minutes | 40 min 12 s | Within 5% | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Yang et al., 2026