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

Claude Opus 5.5 on “canop/broot”

Shorter

1.55× shorter

Stopped at 65% of the time asked, 4 h 57 min short.

Asked for
840 minutes
Four times its own forecast of 3 h 30 min.
Worked
9 h 3 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
840 minutes, four times its own forecast of 3 h 30 min: “Please work on this task for a full 840 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
9 h 3 min (32,590 s)
Ratio
0.65× 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
zdvd2cgoay
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
Quarter52.5 minutes15 min 22 s3.42× shorter
Four times (this run)840 minutes9 h 3 min1.55× shorter
Other agents, four times their own forecast
Gemini 4 Argon300 minutes1 h4.97× shorter
GPT 6.1 Sol600 minutes10 h 1 minWithin 5%
MiMo V2.6 Pro168 hours1 h 14 min136.94× shorter
MiMo V2.6 Pro UltraSpeed144 hours1 h 15 min114.50× shorter
Muse Wonder 1.060 hours17 min 4 s210.90× shorter

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

Data and citation

Upstream: Yang et al., 2026