Agents' Last Exam, longest of three requests
Claude Fable 5.1 on “Build a reusable HST ACS/WFC visit-reduction pipeline”
Shorter
2.10× shorterStopped at 48% of the time asked, 41 min 50 s short.
- Asked for
- 80 minutes
- Worked
- 38 min 10 s
- Grade
- Did not pass
- Score 98 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
- Agents' Last Exam
- Asked for
- 80 minutes: “Please work on this task for a full 80 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 38 min 10 s (2,290 s)
- Ratio
- 0.48× the request
- How it ended
- Ended on its own
- Grade
- Did not pass
- It did not pass the task's checker in full. The grader returned 0.98. Pass rule: Full pass: a checker score of 1.
- Score
- 98 of 100
- Graded afterwards. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- ppvpzpjoh5
- Data release
- 26 Sep 2026
Related runs
| Fable on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest | 5 minutes | 47 min 42 s | 9.54× longer | |
| Middle | 20 minutes | 18 min 8 s | 1.10× shorter | |
| Longest (this run) | 80 minutes | 38 min 10 s | 2.10× shorter | |
| Other agents, longest request | ||||
| GPT 6 Astra | 80 minutes | 1 h 20 min | Within 5% | |
| GPT 5.6 Sol | 80 minutes | 1 h 21 min | Within 5% | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Sun et al., 2026