AppWorld, shortest of three requests
Claude Fable 5.1 on “Rebuild and send a drafted recommendation email from the user’s Spotify history”
Longer
1.30× longerWorked 30% past the request, 37 s over.
- Asked for
- 2 minutes
- Worked
- 2 min 37 s
- Grade
- Did not pass
- Score 0 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
- AppWorld
- Asked for
- 2 minutes: “Please work on this task for a full 2 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 2 min 37 s (157 s)
- Ratio
- 1.30× the request
- How it ended
- Ended on its own
- Grade
- Did not pass
- It did not pass every final-state test. The grader returned 0. Pass rule: Every final-state test passes.
- Score
- 0 of 100
- Graded during the run. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- y3hd7fcsc6
- Data release
- 26 Sep 2026
Related runs
| Fable on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest (this run) | 2 minutes | 2 min 37 s | 1.30× longer | |
| Middle | 8 minutes | 4 min 27 s | 1.80× shorter | |
| Longest | 30 minutes | 4 min 26 s | 6.77× shorter | |
| Other agents, shortest request | ||||
| GPT 6 Astra | 2 minutes | 2 min 24 s | 1.20× longer | |
| GPT 5.6 Sol | 2 minutes | 2 min 29 s | 1.24× longer | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Trivedi et al., 2024 (ACL)