AssistantBench, longest of three requests
Claude Fable 5.1 on “AssistantBench validation task 17”
Shorter
3.38× shorterStopped at 30% of the time asked, 14 min 5 s short.
- Asked for
- 20 minutes
- Worked
- 5 min 55 s
- Grade
- Did not pass
- Score 1.6 of 100. Graded during the run.
- How it ended
- Ended on its own
Every detail of this run
- Agent
- Claude Fable 5.1, in Claude Code
- Benchmark
- AssistantBench
- Asked for
- 20 minutes: “Please work on this task for a full 20 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 5 min 55 s (355 s)
- Ratio
- 0.30× the request
- How it ended
- Ended on its own
- Grade
- Did not pass
- It did not match the reference answer exactly. The grader returned 0.016. Pass rule: Exact match with the reference answer.
- Score
- 1.6 of 100
- Graded during the run. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- d7o7fjhifq
- Data release
- 26 Sep 2026
Related runs
| Fable on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest | 1.25 minutes | 1 min 30 s | 1.19× longer | |
| Middle | 5 minutes | 3 min 29 s | 1.44× shorter | |
| Longest (this run) | 20 minutes | 5 min 55 s | 3.38× shorter | |
| Other agents, longest request | ||||
| GPT 6 Astra | 20 minutes | 20 min 30 s | Within 5% | |
| GPT 5.6 Sol | 20 minutes | 20 min 33 s | Within 5% | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Yoran et al., 2024 (EMNLP)