Task 8 of 16 in OSWorld 2.0

GeoGebra vase and volume

Fable's three runs landed closest to the request (15 min 21 s, 52 min 8 s and 3 h 6 min). Sol's runs landed furthest from the request (35 min 12 s, 1 h 16 min and 4 h 12 min).

Benchmark: OSWorld 2.0; Upstream id 059
Asked for: 15, 60 and 250 min
Runs here: 9 of 9

One task, asked three ways

Astra
Sol
Fable

One line per agent joins its three runs.

What we asked

Please work on this task for a full [15 | 60 | 250] minutes. Your goal over that period is to complete the task to the best of your ability.

One sentence, sent three times with a different time: 15, 60 and 250 min.

The three sentences in full

Shortest15 minutes

Please work on this task for a full 15 minutes. Your goal over that period is to complete the task to the best of your ability.

Middle60 minutes

Please work on this task for a full 60 minutes. Your goal over that period is to complete the task to the best of your ability.

Longest250 minutes

Please work on this task for a full 250 minutes. Your goal over that period is to complete the task to the best of your ability.

Every run on this task

Every run on GeoGebra vase and volume, one row per agent
Agent	Shortest	Middle	Longest
Astra	23 min 53 s 1.59× longer Did not pass, 20 of 100	1 h 1 min Within 5% Did not pass, 20 of 100	4 h 13 min Within 5% Did not pass, 20 of 100
Sol	35 min 12 s 2.35× longer Did not pass, 20 of 100	1 h 16 min 1.26× longer Did not pass, 20 of 100	4 h 12 min Within 5% Did not pass, 20 of 100
Fable	15 min 21 s Within 5% Did not pass, 20 of 100	52 min 8 s 1.15× shorter Did not pass, 20 of 100	3 h 6 min 1.34× shorter Did not pass, 20 of 100

Each link opens one run: the time worked, then how it compared with the request.

Data and citation

Upstream: Yuan et al., 2026