Task 1 of 20 in Humanity's Last Exam
Humanity's Last Exam item 25
Sol 6.1's two runs came closest to the time asked: 46 s when asked for 23 s, and 6 min 16 s when asked for 6 min. MiMo UltraSpeed's two runs landed furthest: 10 s when asked for 3 min 45 s, and 35 s when asked for 1 h.
- Asked for
- 8 s to 1 h
- Runs here
- 12
- 6 agents
Every agent on this task
- Opus 5.5
- Argon
- Sol 6.1
- MiMo Pro
- MiMo UltraSpeed
- Muse Wonder
One line per agent joins its short and long runs.
What we asked
Please work on this task for a full [time]. Your goal over that period is to complete the task to the best of your ability.
Each agent read it twice, with its own short and long ask, as listed below.
Every run on this task
| Agent | Forecast | Runtime | Short ask | Long ask |
|---|---|---|---|---|
| Opus 5.5 | 1 min 30 s | 3 min 43 s | 1 min 21 s 3.58× longer asked 23 s | 8 min 51 s 1.47× longer asked 6 min |
| Argon | 30 s | 1 min 9 s | 1 min 5 s 8.61× longer asked 8 s | 2 min 35 s 1.29× longer asked 2 min |
| Sol 6.1 | 1 min 30 s | 48 s | 46 s 2.06× longer asked 23 s | 6 min 16 s Within 5% asked 6 min |
| MiMo Pro | 3 min | 1 min 13 s | 47 s Within 5% asked 45 s | 2 min 2 s 5.90× shorter asked 12 min |
| MiMo UltraSpeed | 15 min | 10 s | 10 s 23.68× shorter asked 3 min 45 s | 35 s 103.15× shorter asked 1 h |
| Muse Wonder | 2 min | 21 s | 50 s 1.66× longer asked 30 s | 1 min 29 s 5.41× shorter asked 8 min |
No runs match these filters.
Runtime is each agent’s run with no time asked.
Data and citation
Upstream: Phan et al., 2025