WildClawBench, middle of three requests
Claude Fable 5.1 on “Geolocate a street image and return country, city, and coordinates”
Longer
7.10× longerWorked 610% past the request, 30 min 31 s over.
- Asked for
- 5 minutes
- Worked
- 35 min 31 s
- Grade
- Score 0 of 100
- Nothing to grade, counted as 0.
- How it ended
- Ended on its own
Every detail of this run
- Agent
- Claude Fable 5.1, in Claude Code
- Benchmark
- WildClawBench
- Asked for
- 5 minutes: “Please work on this task for a full 5 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 35 min 31 s (2,131 s)
- Ratio
- 7.10× the request
- How it ended
- Ended on its own
- Grade
- 0
- Task grader score, 0 to 1 (a model judge only where the task calls for one). No official pass rule.
- Score
- 0 of 100
- Nothing to grade, counted as 0. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- u7ge6f7dcx
- Data release
- 26 Sep 2026
Related runs
| Fable on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest | 1.25 minutes | 40 min 1 s | 32.02× longer | |
| Middle (this run) | 5 minutes | 35 min 31 s | 7.10× longer | |
| Longest | 20 minutes | 30 min 36 s | 1.53× longer | |
| Other agents, middle request | ||||
| GPT 6 Astra | 5 minutes | 40 min | 8.00× longer | |
| GPT 5.6 Sol | 5 minutes | 40 min | 8.00× longer | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Ding et al., 2026