OSWorld 2.0, longest of three requests

GPT 5.6 Sol on “KiCad schematic and PCB”

Longer

1.38× longer

Worked 38% past the request, 1 h 35 min over.

Asked

Worked

Dashed line: safety cutoff at 8 h 20 min

Every detail of this run

Agent: GPT 5.6 Sol, in Codex
Benchmark: OSWorld 2.0
Task: KiCad schematic and PCB
Asked for: 250 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.”
Worked: 5 h 45 min (20,682 s)
Ratio: 1.38× the request
How it ended: Ended on its own
Grade: Did not pass; It did not meet every checkpoint. The grader returned 0.875. Pass rule: Binary completion: a score of 1.00, with every checkpoint met.
Score: 87.5 of 100; Graded during the run. The paper's value.
Wording: The standard sentence, used from 4 September 2026.
Run id: tjupp7vqvl
Data release: 26 Sep 2026

The other requests for GPT 5.6 Sol on this task, and the other agents at this request
Sol on this task	Asked	Worked	Result
Shortest	15 minutes	2 h 4 min	8.26× longer
Middle	60 minutes	1 h 3 min	1.05× longer
Longest (this run)	250 minutes	5 h 45 min	1.38× longer
Other agents, longest request
GPT 6 Astra	250 minutes	4 h 11 min	Within 5%
Claude Fable 5.1	250 minutes	4 h 2 min	Within 5%

See it on the chart. Transcripts are not published yet.

Data and citation