PPTArena, longest of three requests
GPT 6 Astra on “Swimlane process diagram”
Within 5%
Within 5%Worked 101% of the time asked, within 5% of it.
- Asked for
- 25 minutes
- Worked
- 25 min 14 s
- Grade
- Score 80 of 100
- Instructions 3 of 5, visuals 5 of 5. Graded during the run.
- How it ended
- Ended on its own
Every detail of this run
- Agent
- GPT 6 Astra, in Codex
- Benchmark
- PPTArena
- Asked for
- 25 minutes: “Please work on this task for a full 25 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 25 min 14 s (1,514 s)
- Ratio
- 1.01× the request
- How it ended
- Ended on its own
- Grade
- Instructions 3.00 of 5, visuals 5.00 of 5
- Two 0-to-5 judge scores, instruction following (IF) and visual quality (VQ), each the median of 3 judge samples. No official pass rule.
- Score
- 80 of 100
- Graded during the run. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- dm6zuk6zt4
- Data release
- 26 Sep 2026
Related runs
| Astra on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest | 1.5 minutes | 1 min 44 s | 1.16× longer | |
| Middle | 6 minutes | 6 min 20 s | 1.05× longer | |
| Longest (this run) | 25 minutes | 25 min 14 s | Within 5% | |
| Other agents, longest request | ||||
| GPT 5.6 Sol | 25 minutes | 26 min 45 s | 1.07× longer | |
| Claude Fable 5.1 | 25 minutes | 15 min 52 s | 1.58× shorter | |
See it on the chart. Transcripts are not published yet.