TUA-Bench, shortest of three requests

GPT 5.6 Sol on “APA references review”

Longer

1.45× longer

Worked 45% past the request, 1 min 7 s over.

Asked

Worked

Safety cutoff 1 h 20 min, off the scale

Every detail of this run

Agent: GPT 5.6 Sol, in Codex
Benchmark: TUA-Bench
Task: APA references review
Asked for: 2.5 minutes: “Please work on this task for a full 2.5 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked: 3 min 37 s (217 s)
Ratio: 1.45× the request
How it ended: Ending not labelled yet
Grade: Did not pass; It did not complete the task in full. The grader returned 0.657. Pass rule: Full completion: a reward of 1 on the final state.
Score: 65.7 of 100; Graded during the run. The paper's value.
Wording: The standard sentence, used from 4 September 2026.
Run id: sq5ioej5r2
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 (this run)	2.5 minutes	3 min 37 s	1.45× longer
Middle	10 minutes	10 min 49 s	1.08× longer
Longest	40 minutes	41 min 5 s	Within 5%
Other agents, shortest request
GPT 6 Astra	2.5 minutes	2 min 45 s	1.10× longer
Claude Fable 5.1	2.5 minutes	9 min 20 s	3.73× longer
Muse Spark 1.3	2.5 minutes	2 min 9 s	1.16× shorter

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

Read before quoting. How this run ended has not been labelled yet. It is timed like every other run and counted in every number. Details

Data and citation