TUA-Bench, shortest of three requests
Claude Fable 5.1 on “Optimize cold plate”
Longer
2.96× longerWorked 196% past the request, 11 min 47 s over.
- Asked for
- 6 minutes
- Worked
- 17 min 47 s
- Grade
- Did not pass
- Score 33 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
- TUA-Bench
- Asked for
- 6 minutes: “Please work on this task for a full 6 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 17 min 47 s (1,067 s)
- Ratio
- 2.96× the request
- How it ended
- Ended on its own
- Grade
- Did not pass
- It did not complete the task in full. The grader returned 0.33. Pass rule: Full completion: a reward of 1 on the final state.
- Score
- 33 of 100
- Graded during the run. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- 4dszdxyjiy
- Data release
- 26 Sep 2026
Related runs
| Fable on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest (this run) | 6 minutes | 17 min 47 s | 2.96× longer | |
| Middle | 25 minutes | 28 min 9 s | 1.13× longer | |
| Longest | 100 minutes | 35 min 45 s | 2.80× shorter | |
| Other agents, shortest request | ||||
| GPT 6 Astra | 6 minutes | 6 min 30 s | 1.08× longer | |
| GPT 5.6 Sol | 6 minutes | 9 min 55 s | 1.65× longer | |
| Muse Spark 1.3 | 6 minutes | 15 min 2 s | 2.51× longer | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Chen et al., 2026