ProgramBench, shortest of three requests

GPT 5.6 Sol on “stathissideris/ditaa”

Longer

2.69× longer

Worked 169% past the request, 13 min 29 s over.

Asked for
8 minutes
Worked
21 min 29 s
Grade
Did not pass
Score 19.7 of 100. Graded afterwards.
How it ended
Ending not labelled yet
Every detail of this run
Agent
GPT 5.6 Sol, in Codex
Benchmark
ProgramBench
Asked for
8 minutes: “Please work on this task for a full 8 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
21 min 29 s (1,289 s)
Ratio
2.69× the request
How it ended
Ending not labelled yet
Grade
Did not pass
It did not pass every hidden test. The grader returned 0.197. Pass rule: Resolved: every hidden test passes.
Score
19.7 of 100
Graded afterwards. The paper's value.
Wording
The standard sentence, used from 4 September 2026.
Run id
cpjmmnj4ps
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 taskAskedWorkedSplitResult
Shortest (this run)8 minutes21 min 29 s2.69× longer
Middle30 minutes28 min 32 sWithin 5%
Longest125 minutes2 h 6 minWithin 5%
Other agents, shortest request
GPT 6 Astra8 minutes8 min 12 sWithin 5%
Claude Fable 5.18 minutes5 min 16 s1.52× 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

Upstream: Yang et al., 2026