Task 12 of 12 in Agents' Last Exam
Molecular Structure Plausibility
Sol 6.1's two runs came closest to the time asked: 3 min 17 s when asked for 3 min, and 48 min 37 s when asked for 48 min. Muse Wonder worked about the same whatever it was asked: 1 min 50 s when asked for 8 min 45 s, and 2 min 17 s when asked for 2 h 20 min.
- Asked for
- 2 min to 2 h 40 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 | 15 min | 44 min 4 s | 1 min 52 s 2.02× shorter asked 3 min 45 s | 28 min 27 s 2.11× shorter asked 1 h |
| Argon | 8 min | 32 min 52 s | 14 min 56 s 7.46× longer asked 2 min | 33 min 50 s 1.06× longer asked 32 min |
| Sol 6.1 | 12 min | 4 min 26 s | 3 min 17 s 1.09× longer asked 3 min | 48 min 37 s Within 5% asked 48 min |
| MiMo Pro | 35 min | 13 min 31 s | 14 min 45 s 1.69× longer asked 8 min 45 s | 18 min 16 s 7.67× shorter asked 2 h 20 min |
| MiMo UltraSpeed | 40 min | 4 min 5 s | 4 min 11 s 2.39× shorter asked 10 min | 4 min 7 s 38.93× shorter asked 2 h 40 min |
| Muse Wonder | 35 min | 1 min 20 s | 1 min 50 s 4.76× shorter asked 8 min 45 s | 2 min 17 s 61.40× shorter asked 2 h 20 min |
No runs match these filters.
Runtime is each agent’s run with no time asked.
Data and citation
Upstream: Sun et al., 2026