Humanity's Last Exam, AgentTime v1.1, asked for four times its own forecast
Claude Opus 5.5 on “Humanity's Last Exam item 30”
Within 5%
Within 5%Worked 97% of the time asked, within 5% of it.
- Asked for
- 68 minutes
- Four times its own forecast of 17 min.
- Worked
- 1 h 6 min
- Grade
- Not graded
- How it ended
- Ended on its own
Every detail of this run
- Agent
- Claude Opus 5.5, in Claude Code
- Benchmark
- Humanity's Last Exam
- Asked for
- 68 minutes, four times its own forecast of 17 min: “Please work on this task for a full 68 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 1 h 6 min (3,973 s)
- Ratio
- 0.97× 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
- gsi7ikmcnl
- Data release
- 2 Nov 2026
Related runs
| Opus 5.5 on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter | 4.25 minutes | 4 min 44 s | 1.11× longer | |
| Four times (this run) | 68 minutes | 1 h 6 min | Within 5% | |
| Other agents, four times their own forecast | ||||
| Gemini 4 Argon | 840 seconds | 14 min 36 s | Within 5% | |
| GPT 6.1 Sol | 28 minutes | 28 min 21 s | Within 5% | |
| MiMo V2.6 Pro | 120 minutes | 1 h 8 min | 1.76× shorter | |
| MiMo V2.6 Pro UltraSpeed | 140 minutes | 12 min 18 s | 11.39× shorter | |
| Muse Wonder 1.0 | 88 minutes | 15 min 22 s | 5.73× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Phan et al., 2025