Task 9 of 12 in BrowseComp
BrowseComp question 9
Muse Wonder's two runs followed the time asked: 4 min 35 s when asked for 4 min 30 s, and 1 h 12 min when asked for 1 h 12 min. MiMo UltraSpeed worked about the same whatever it was asked: 6 min 24 s when asked for 8 min 45 s, and 3 min 28 s when asked for 2 h 20 min.
- Asked for
- 45 s to 2 h 20 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 | 12 min | 5 min 39 s | 7 min 55 s 2.64× longer asked 3 min | 49 min 31 s Within 5% asked 48 min |
| Argon | 3 min | 3 min 8 s | 1 min 26 s 1.90× longer asked 45 s | 13 min 15 s 1.10× longer asked 12 min |
| Sol 6.1 | 12 min | 4 min 2 s | 3 min 21 s 1.11× longer asked 3 min | 48 min 59 s Within 5% asked 48 min |
| MiMo Pro | 25 min | 31 min 56 s | 47 min 59 s 7.68× longer asked 6 min 15 s | 19 min 24 s 5.15× shorter asked 1 h 40 min |
| MiMo UltraSpeed | 35 min | 5 min 3 s | 6 min 24 s 1.37× shorter asked 8 min 45 s | 3 min 28 s 40.48× shorter asked 2 h 20 min |
| Muse Wonder | 18 min | 4 min 27 s | 4 min 35 s Within 5% asked 4 min 30 s | 1 h 12 min Within 5% asked 1 h 12 min |
No runs match these filters.
Runtime is each agent’s run with no time asked.
Data and citation
Upstream: Wei et al., 2025