Task 12 of 18 in OSWorld
Radio bumper
Sol 6.1's two runs followed the time asked: 9 min 31 s when asked for 8 min 45 s, and 2 h 22 min when asked for 2 h 20 min. Argon's two runs landed furthest: 1 h 11 min when asked for 3 min 45 s, and 10 min 16 s when asked for 1 h.
- Asked for
- 3 min 45 s to 4 h 10 min
- Runs here
- 16
- 8 agents
Every agent on this task
- Astra
- Sol
- Fable
- Opus 5.5
- Argon
- Sol 6.1
- MiMo Pro
- MiMo UltraSpeed
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 |
|---|---|---|---|---|
| Astra | 30 min | 44 min 52 s | 22 min 29 s 1.50× longer asked 15 min | 4 h 11 min Within 5% asked 4 h 10 min |
| Sol | 25 min | 1 h 13 min | 1 h 16 min 5.08× longer asked 15 min | 1 h 2 min 4.02× shorter asked 4 h 10 min |
| Fable | 45 min | 33 min 3 s | 27 min 10 s 1.81× longer asked 15 min | 4 h 12 min Within 5% asked 4 h 10 min |
| Opus 5.5 | 32 min | 42 min 41 s | 31 min 56 s 3.99× longer asked 8 min | 1 h 44 min 1.23× shorter asked 2 h 8 min |
| Argon | 15 min | 1 h 18 min | 1 h 11 min 19.05× longer asked 3 min 45 s | 10 min 16 s 5.85× shorter asked 1 h |
| Sol 6.1 | 35 min | 18 min 38 s | 9 min 31 s 1.09× longer asked 8 min 45 s | 2 h 22 min Within 5% asked 2 h 20 min |
| MiMo Pro | 32 min | 1 h 20 min | 1 h 16 min 9.51× longer asked 8 min | 2 h 7 min Within 5% asked 2 h 8 min |
| MiMo UltraSpeed | 55 min | 41 min 57 s | 59 min 19 s 4.31× longer asked 13 min 45 s | 24 min 45 s 8.89× shorter asked 3 h 40 min |
No runs match these filters.
Runtime is a v1.1 agent’s run with no time asked, and for the paper’s three agents the paper’s estimate from their timed runs.
Data and citation
Upstream: Yuan et al., 2026