Humanity's Last Exam, AgentTime v1.1, asked for four times its own forecast
Muse Wonder 1.0 on “Humanity's Last Exam item 42”
Shorter
1.89× shorterStopped at 53% of the time asked, 3 min 46 s short.
- Asked for
- 480 seconds
- Four times its own forecast of 2 min.
- Worked
- 4 min 14 s
- Grade
- Not graded
- How it ended
- Ended on its own
Every detail of this run
- Agent
- Muse Wonder 1.0, in Muse Code
- Benchmark
- Humanity's Last Exam
- Asked for
- 480 seconds, four times its own forecast of 2 min: “Please work on this task for a full 480 seconds. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 4 min 14 s (254 s)
- Ratio
- 0.53× 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
- 7zy6wpqbc6
- Data release
- 2 Nov 2026
Related runs
| Muse Wonder on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter | 30 seconds | 50 s | 1.65× longer | |
| Four times (this run) | 480 seconds | 4 min 14 s | 1.89× shorter | |
| Other agents, four times their own forecast | ||||
| Claude Opus 5.5 | 240 seconds | 5 min 1 s | 1.26× longer | |
| Gemini 4 Argon | 168 seconds | 3 min 6 s | 1.11× longer | |
| GPT 6.1 Sol | 480 seconds | 8 min 10 s | Within 5% | |
| MiMo V2.6 Pro | 16 minutes | 31 min 21 s | 1.96× longer | |
| MiMo V2.6 Pro UltraSpeed | 16 minutes | 2 min 29 s | 6.43× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Phan et al., 2025