Task 1 of 12 in BrowseComp
BrowseComp question 1
Sol 6.1's two runs followed the time asked: 6 min 26 s when asked for 6 min 15 s, and 1 h 41 min when asked for 1 h 40 min. MiMo UltraSpeed worked about the same whatever it was asked: 5 min 53 s when asked for 11 min 15 s, and 10 min 27 s when asked for 3 h.
- Asked for
- 3 min 30 s to 5 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 | 4 h | 49 min 47 s 6.64× longer asked 7 min 30 s | 1 h 9 min 1.73× shorter asked 2 h |
| Argon | 14 min | 4 h 32 min | 4 min 9 s 1.18× longer asked 3 min 30 s | 56 min 14 s Within 5% asked 56 min |
| Sol 6.1 | 25 min | 57 min 37 s | 6 min 26 s Within 5% asked 6 min 15 s | 1 h 41 min Within 5% asked 1 h 40 min |
| MiMo Pro | 1 h 15 min | 1 h 43 min | 1 h 2 min 3.31× longer asked 18 min 45 s | 1 h 32 min 3.26× shorter asked 5 h |
| MiMo UltraSpeed | 45 min | 16 min 35 s | 5 min 53 s 1.91× shorter asked 11 min 15 s | 10 min 27 s 17.23× shorter asked 3 h |
| Muse Wonder | 40 min | 51 min 8 s | 19 min 37 s 1.96× longer asked 10 min | 2 h 31 min 1.06× shorter asked 2 h 40 min |
No runs match these filters.
Runtime is each agent’s run with no time asked.
Data and citation
Upstream: Wei et al., 2025