Task 11 of 18 in OSWorld
JSFX and REAPER render
Sol 6.1's two runs came closest to the time asked: 13 min 22 s when asked for 9 min 30 s, and 2 h 34 min when asked for 2 h 32 min. Fable's two runs landed furthest: 8 h 30 min when asked for 15 min, and 44 min 1 s when asked for 4 h 10 min.
- Asked for
- 4 min 30 s to 7 h
- Runs here
- 16
- See Read before quoting below
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 | 49 min 54 s | 28 min 12 s 1.88× longer asked 15 min | 4 h 11 min Within 5% asked 4 h 10 min |
| Sol | 25 min | 26 min 55 s | 33 min 56 s 2.26× longer asked 15 min | 26 min 7 s 9.57× shorter asked 4 h 10 min |
| Fable | 45 min | 29 min 19 s | 8 h 30 min 34.00× longer asked 15 min | 44 min 1 s 5.68× shorter asked 4 h 10 min |
| Opus 5.5 | 40 min | 1 h 30 min | 40 min 55 s 4.09× longer asked 10 min | 1 h 4 min 2.50× shorter asked 2 h 40 min |
| Argon | 18 min | 1 h 21 min | 14 min 39 s 3.26× longer asked 4 min 30 s | 1 h 37 min 1.35× longer asked 1 h 12 min |
| Sol 6.1 | 38 min | 22 min 13 s | 13 min 22 s 1.41× longer asked 9 min 30 s | 2 h 34 min Within 5% asked 2 h 32 min |
| MiMo Pro | 1 h 10 min | 2 h 7 min | 1 h 3.44× longer asked 17 min 30 s | 3 h 18 min 1.41× shorter asked 4 h 40 min |
| MiMo UltraSpeed | 1 h 45 min | 1 h 13 min | 53 min 43 s 2.05× longer asked 26 min 15 s | 1 h 31 min 4.61× shorter asked 7 h |
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.
Read before quoting. 1 run was stopped by AgentTime, so its time is a lower bound. Details
Data and citation
Upstream: Yuan et al., 2026