Task 17 of 20 in Humanity's Last Exam
Humanity's Last Exam item 41
Sol 6.1's two runs came closest to the time asked: 20 s when asked for 8 s, and 2 min 11 s when asked for 2 min. MiMo Pro's two runs landed furthest: 3 min 44 s when asked for 30 s, and 15 min 25 s when asked for 8 min.
- Asked for
- 5 s to 8 min
- 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 | 24 s | 35 s | 35 s 5.82× longer asked 6 s | 2 min 27 s 1.53× longer asked 1 min 36 s |
| Argon | 18 s | 24 s | 32 s 7.07× longer asked 5 s | 1 min 20 s 1.11× longer asked 1 min 12 s |
| Sol 6.1 | 30 s | 11 s | 20 s 2.61× longer asked 8 s | 2 min 11 s 1.09× longer asked 2 min |
| MiMo Pro | 2 min | 26 s | 3 min 44 s 7.45× longer asked 30 s | 15 min 25 s 1.93× longer asked 8 min |
| MiMo UltraSpeed | 1 min | 14 s | 27 s 1.79× longer asked 15 s | 44 s 5.47× shorter asked 4 min |
| Muse Wonder | 30 s | 21 s | 46 s 6.07× longer asked 8 s | 2 min 14 s 1.12× longer asked 2 min |
No runs match these filters.
Runtime is each agent’s run with no time asked.
Data and citation
Upstream: Phan et al., 2025