TUA-Bench, shortest of three requests

GPT 5.6 Sol on “Optimize cold plate”

Longer

1.65× longer

Worked 65% past the request, 3 min 55 s over.

Asked for
6 minutes
Worked
9 min 55 s
Grade
Did not pass
Score 33 of 100. Graded during the run.
How it ended
Ended on its own
Every detail of this run
Agent
GPT 5.6 Sol, in Codex
Benchmark
TUA-Bench
Asked for
6 minutes: “Please work on this task for a full 6 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
9 min 55 s (595 s)
Ratio
1.65× the request
How it ended
Ended on its own
Grade
Did not pass
It did not complete the task in full. The grader returned 0.33. Pass rule: Full completion: a reward of 1 on the final state.
Score
33 of 100
Graded during the run. The paper's value.
Wording
The standard sentence, used from 4 September 2026.
Run id
tps4ole7n4
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)6 minutes9 min 55 s1.65× longer
Middle25 minutes25 min 59 sWithin 5%
Longest100 minutes1 h 41 minWithin 5%
Other agents, shortest request
GPT 6 Astra6 minutes6 min 30 s1.08× longer
Claude Fable 5.16 minutes17 min 47 s2.96× longer
Muse Spark 1.36 minutes15 min 2 s2.51× longer

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

Data and citation

Upstream: Chen et al., 2026