ProgramBench, AgentTime v1.1, asked for four times its own forecast

Claude Opus 5.5 on “tomarrell/wrapcheck”

Shorter

2.14× shorter

Stopped at 47% of the time asked, 2 h 40 min short.

Asked for
300 minutes
Four times its own forecast of 1 h 15 min.
Worked
2 h 20 min
Grade
Not graded
How it ended
Ended on its own
Every detail of this run
Agent
Claude Opus 5.5, in Claude Code
Benchmark
ProgramBench
Asked for
300 minutes, four times its own forecast of 1 h 15 min: “Please work on this task for a full 300 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
2 h 20 min (8,398 s)
Ratio
0.47× the request
How it ended
Ended on its own
Grade
Not graded
Clock tool
The agent could call a wall-clock tool to read the current time.
Wording
The paper's standard sentence, sent in AgentTime v1.1.
Run id
73pt3w3gbk
Data release
2 Nov 2026
The other ask for Claude Opus 5.5 on this task, and the other agents at the same share of their own time
Opus 5.5 on this taskAskedWorkedSplitResult
Quarter18.75 minutes5 min 4 s3.71× shorter
Four times (this run)300 minutes2 h 20 min2.14× shorter
Other agents, four times their own forecast
Gemini 4 Argon168 minutes3 min 24 s49.41× shorter
GPT 6.1 Sol240 minutes4 h 1 minWithin 5%
MiMo V2.6 Pro24 hours1 h 26 min16.66× shorter
MiMo V2.6 Pro UltraSpeed28 hours53 min 27 s31.43× shorter
Muse Wonder 1.020 hours9 min 50 s122.08× shorter

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

Data and citation

Upstream: Yang et al., 2026