Task 10 of 12 in BrowseComp
BrowseComp question 10
Sol 6.1's two runs followed the time asked: 6 min 29 s when asked for 6 min, and 1 h 37 min when asked for 1 h 36 min. MiMo UltraSpeed's two runs landed furthest: 4 min 45 s when asked for 7 min 30 s, and 15 min 32 s when asked for 2 h.
- Asked for
- 3 min to 4 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 | 30 min | 23 min 23 s | 43 min 35 s 5.81× longer asked 7 min 30 s | 2 h Within 5% asked 2 h |
| Argon | 12 min | 15 min | 9 min 36 s 3.20× longer asked 3 min | 48 min 19 s Within 5% asked 48 min |
| Sol 6.1 | 24 min | 2 min 47 s | 6 min 29 s 1.08× longer asked 6 min | 1 h 37 min Within 5% asked 1 h 36 min |
| MiMo Pro | 1 h | 40 min 55 s | 36 min 52 s 2.46× longer asked 15 min | 1 h 59 min 2.01× shorter asked 4 h |
| MiMo UltraSpeed | 30 min | 5 min 55 s | 4 min 45 s 1.58× shorter asked 7 min 30 s | 15 min 32 s 7.72× shorter asked 2 h |
| Muse Wonder | 12 min | 8 min 38 s | 4 min 39 s 1.55× longer asked 3 min | 49 min 40 s Within 5% asked 48 min |
No runs match these filters.
Runtime is each agent’s run with no time asked.
Data and citation
Upstream: Wei et al., 2025