AssistantBench, middle of three requests
GPT 5.6 Sol on “AssistantBench validation task 18”
Longer
1.94× longerWorked 94% past the request, 4 min 43 s over.
- Asked for
- 5 minutes
- Worked
- 9 min 43 s
- Grade
- Did not pass
- Score 10.7 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
- 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
- 9 min 43 s (583 s)
- Ratio
- 1.94× the request
- How it ended
- Ended on its own
- Grade
- Did not pass
- It did not match the reference answer exactly. The grader returned 0.107. Pass rule: Exact match with the reference answer.
- Score
- 10.7 of 100
- Graded afterwards. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- tyjgjp2eky
- Data release
- 26 Sep 2026
Related runs
| Sol on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest | 1.25 minutes | 9 min 42 s | 7.77× longer | |
| Middle (this run) | 5 minutes | 9 min 43 s | 1.94× longer | |
| Longest | 20 minutes | 14 min 12 s | 1.41× shorter | |
| Other agents, middle request | ||||
| GPT 6 Astra | 5 minutes | 6 min 21 s | 1.27× longer | |
| Claude Fable 5.1 | 5 minutes | 8 min 13 s | 1.64× longer | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Yoran et al., 2024 (EMNLP)