Task 12 of 20 in Humanity's Last Exam
Humanity's Last Exam item 36
Sol 6.1's two runs came closest to the time asked: 1 min 56 s when asked for 1 min 15 s, and 20 min 21 s when asked for 20 min. MiMo Pro worked about the same whatever it was asked: 28 min 37 s when asked for 3 min, and 49 min 53 s when asked for 48 min.
- Asked for
- 23 s to 48 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 | 1 min 30 s | 21 s | 43 s 1.91× longer asked 23 s | 6 min 23 s 1.06× longer asked 6 min |
| Argon | 1 min 30 s | 2 min 3 s | 1 min 48 s 4.79× longer asked 23 s | 6 min 35 s 1.10× longer asked 6 min |
| Sol 6.1 | 5 min | 6 min 21 s | 1 min 56 s 1.55× longer asked 1 min 15 s | 20 min 21 s Within 5% asked 20 min |
| MiMo Pro | 12 min | 37 min 8 s | 28 min 37 s 9.54× longer asked 3 min | 49 min 53 s Within 5% asked 48 min |
| MiMo UltraSpeed | 12 min | 9 min 39 s | 2 min 16 s 1.33× shorter asked 3 min | 10 min 51 s 4.42× shorter asked 48 min |
| Muse Wonder | 5 min | 7 min 43 s | 2 min 50 s 2.27× longer asked 1 min 15 s | 11 min 42 s 1.71× shorter asked 20 min |
No runs match these filters.
Runtime is each agent’s run with no time asked.
Data and citation
Upstream: Phan et al., 2025