GPQA Diamond, AgentTime v1.1, asked for four times its own forecast
GPT 6.1 Sol on “GPQA Diamond closed-book question 8 (Chemistry)”
Within 5%
Within 5%Worked 103% of the time asked, within 5% of it.
- Asked for
- 360 seconds
- Four times its own forecast of 1 min 30 s.
- Worked
- 6 min 10 s
- Grade
- Not graded
- How it ended
- Ended on its own
Every detail of this run
- Agent
- GPT 6.1 Sol, in Codex
- Benchmark
- GPQA Diamond
- Asked for
- 360 seconds, four times its own forecast of 1 min 30 s: “Please work on this task for a full 360 seconds. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 6 min 10 s (370 s)
- Ratio
- 1.03× 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
- xi73w4lsmd
- Data release
- 2 Nov 2026
Related runs
| Sol 6.1 on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter | 22.5 seconds | 36 s | 1.58× longer | |
| Four times (this run) | 360 seconds | 6 min 10 s | Within 5% | |
| Other agents, four times their own forecast | ||||
| Claude Opus 5.5 | 360 seconds | 7 min 5 s | 1.18× longer | |
| Gemini 4 Argon | 120 seconds | 2 min 28 s | 1.23× longer | |
| MiMo V2.6 Pro | 20 minutes | 23 min 34 s | 1.18× longer | |
| MiMo V2.6 Pro UltraSpeed | 20 minutes | 3 min 58 s | 5.04× shorter | |
| Muse Wonder 1.0 | 720 seconds | 1 min 47 s | 6.76× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Rein et al., 2024 (COLM)