TUA-Bench, longest of three requests

GPT 5.6 Sol on “Extract gym auditorium”

Longer

1.08× longer

Worked 8% past the request, 1 min 32 s over.

Asked

Worked

Safety cutoff 40 min, off the scale

Every detail of this run

Agent: GPT 5.6 Sol, in Codex
Benchmark: TUA-Bench
Task: Extract gym auditorium
Asked for: 20 minutes: “Please work on this task for a full 20 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked: 21 min 32 s (1,292 s)
Ratio: 1.08× the request
How it ended: Ended on its own
Grade: Passed; It completed the task in full. The grader returned 1. Pass rule: Full completion: a reward of 1 on the final state.
Score: 100 of 100; Graded during the run. The paper's value.
Wording: The standard sentence, used from 4 September 2026.
Run id: yabp5fcz3k
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	1.25 minutes	7 min 6 s	5.68× longer
Middle	5 minutes	6 min 20 s	1.27× longer
Longest (this run)	20 minutes	21 min 32 s	1.08× longer
Other agents, longest request
GPT 6 Astra	20 minutes	20 min 27 s	Within 5%
Claude Fable 5.1	20 minutes	22 min 12 s	1.11× longer
Muse Spark 1.3	20 minutes	8 min 15 s	2.42× shorter

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

Data and citation