Humanity's Last Exam, AgentTime v1.1, asked for four times its own forecast
Muse Wonder 1.0 on “Humanity's Last Exam item 41”
Longer
1.12× longerWorked 12% past the request, 14 s over.
- Asked for
- 120 seconds
- Four times its own forecast of 30 s.
- Worked
- 2 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
- 120 seconds, four times its own forecast of 30 s: “Please work on this task for a full 120 seconds. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 2 min 14 s (134 s)
- Ratio
- 1.12× 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
- tk3pbcptpo
- Data release
- 2 Nov 2026
Related runs
| Muse Wonder on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter | 7.5 seconds | 46 s | 6.07× longer | |
| Four times (this run) | 120 seconds | 2 min 14 s | 1.12× longer | |
| Other agents, four times their own forecast | ||||
| Claude Opus 5.5 | 96 seconds | 2 min 27 s | 1.53× longer | |
| Gemini 4 Argon | 72 seconds | 1 min 20 s | 1.11× longer | |
| GPT 6.1 Sol | 120 seconds | 2 min 11 s | 1.09× longer | |
| MiMo V2.6 Pro | 480 seconds | 15 min 25 s | 1.93× longer | |
| MiMo V2.6 Pro UltraSpeed | 240 seconds | 44 s | 5.47× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Phan et al., 2025