TUA-Bench, AgentTime v1.1, asked for four times its own forecast
GPT 6.1 Sol on “APA references review”
Within 5%
Within 5%Worked 102% of the time asked, within 5% of it.
- Asked for
- 36 minutes
- Four times its own forecast of 9 min.
- Worked
- 36 min 53 s
- Grade
- Not graded
- How it ended
- Ended on its own
Every detail of this run
- Agent
- GPT 6.1 Sol, in Codex
- Benchmark
- TUA-Bench
- Asked for
- 36 minutes, four times its own forecast of 9 min: “Please work on this task for a full 36 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 36 min 53 s (2,213 s)
- Ratio
- 1.02× the request
- How it ended
- Ended on its own
- Grade
- Not graded
- Clock tool
- The agent could call a wall-clock tool to read the current time.
- Wording
- The paper's standard sentence, sent in AgentTime v1.1.
- Run id
- utaka4ovn3
- Data release
- 2 Nov 2026
Related runs
| Sol 6.1 on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter | 2.25 minutes | 2 min 30 s | 1.11× longer | |
| Four times (this run) | 36 minutes | 36 min 53 s | Within 5% | |
| Other agents, four times their own forecast | ||||
| Claude Opus 5.5 | 64 minutes | 15 min 10 s | 4.22× shorter | |
| Gemini 4 Argon | 16 minutes | 18 min 48 s | 1.18× longer | |
| MiMo V2.6 Pro | 200 minutes | 12 min 7 s | 16.50× shorter | |
| MiMo V2.6 Pro UltraSpeed | 160 minutes | 7 min 40 s | 20.86× shorter | |
| Muse Wonder 1.0 | 96 minutes | 58 s | 98.97× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Chen et al., 2026