AssistantBench, shortest of three requests
GPT 5.6 Sol on “AssistantBench validation task 4”
Longer
2.31× longerWorked 131% past the request, 1 min 39 s over.
- Asked for
- 1.25 minutes
- Worked
- 2 min 54 s
- Grade
- Did not pass
- Score 0 of 100. Graded afterwards.
- How it ended
- Ended on its own
Every detail of this run
- Agent
- GPT 5.6 Sol, in Codex
- Benchmark
- AssistantBench
- Asked for
- 1.25 minutes: “Please work on this task for a full 1.25 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 2 min 54 s (174 s)
- Ratio
- 2.31× 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. Pass rule: Exact match with the reference answer.
- Score
- 0 of 100
- Graded afterwards. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- xqwagnldqs
- Data release
- 26 Sep 2026
Related runs
| Sol on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest (this run) | 1.25 minutes | 2 min 54 s | 2.31× longer | |
| Middle | 5 minutes | 5 min 13 s | Within 5% | |
| Longest | 20 minutes | 2 min 30 s | 7.99× shorter | |
| Other agents, shortest request | ||||
| GPT 6 Astra | 1.25 minutes | 1 min 29 s | 1.18× longer | |
| Claude Fable 5.1 | 1.25 minutes | 1 min | 1.24× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Yoran et al., 2024 (EMNLP)