AssistantBench, longest of three requests

GPT 5.6 Sol on “AssistantBench validation task 5”

Shorter

1.66× shorter

Stopped at 60% of the time asked, 7 min 57 s short.

Asked

Worked

Safety cutoff 40 min, off the scale

Every detail of this run

Agent: GPT 5.6 Sol, in Codex
Benchmark: AssistantBench
Task: AssistantBench validation task 5
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: 12 min 3 s (723 s)
Ratio: 0.60× the request
How it ended: Ended on its own
Grade: Did not pass; It did not match the reference answer exactly. The grader returned 0. Pass rule: Exact match with the reference answer.
Score: 0 of 100; Graded afterwards. The paper's value.
Wording: The standard sentence, used from 4 September 2026.
Run id: fyobioda27
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	3 min 38 s	2.91× longer
Middle	5 minutes	2 min 11 s	2.29× shorter
Longest (this run)	20 minutes	12 min 3 s	1.66× shorter
Other agents, longest request
GPT 6 Astra	20 minutes	9 min 37 s	2.08× shorter
Claude Fable 5.1	20 minutes	40 min	2.00× longer

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

Data and citation