TUA-Bench, longest of three requests
GPT 6 Astra on “Daily email report”
Within 5%
Within 5%Worked 101% of the time asked, within 5% of it.
- Asked for
- 25 minutes
- Worked
- 25 min 15 s
- Grade
- Passed
- Score 100 of 100. Graded during the run.
- How it ended
- Ended on its own
Every detail of this run
- Agent
- GPT 6 Astra, in Codex
- Benchmark
- TUA-Bench
- 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 15 s (1,515 s)
- Ratio
- 1.01× the request
- How it ended
- Ended on its own
- Grade
- Passed
- It completed the task in full. The grader returned 1. Pass rule: Full completion: a reward of 1 on the final state.
- Score
- 100 of 100
- Graded during the run. The paper's value.
- Wording
- The standard sentence, used from 4 September 2026.
- Run id
- e6zexdmd5i
- Data release
- 26 Sep 2026
Related runs
| Astra on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Shortest | 1.5 minutes | 1 min 43 s | 1.15× longer | |
| Middle | 6 minutes | 6 min 9 s | Within 5% | |
| Longest (this run) | 25 minutes | 25 min 15 s | Within 5% | |
| Other agents, longest request | ||||
| GPT 5.6 Sol | 25 minutes | 25 min 40 s | Within 5% | |
| Claude Fable 5.1 | 25 minutes | 2 min 9 s | 11.67× shorter | |
| Muse Spark 1.3 | 25 minutes | 2 min 10 s | 11.54× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Chen et al., 2026