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

Claude Opus 5.5 on “luajit/luajit”

Longer

1.75× longer

Worked 75% past the request, 12 h 54 min over.

Asked for
17.3 hours
Four times its own forecast of 4 h 20 min.
Worked
30 h 12 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
17.3 hours, four times its own forecast of 4 h 20 min: “Please work on this task for a full 17.3 hours. Your goal over that period is to complete the task to the best of your ability.”
Worked
30 h 12 min (108,733 s)
Ratio
1.75× 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
uer3shxpaq
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
Quarter1.08 hours52 min 10 s1.24× shorter
Four times (this run)17.3 hours30 h 12 min1.75× longer
Other agents, four times their own forecast
Gemini 4 Argon380 minutes1 h 49 min3.48× shorter
GPT 6.1 Sol600 minutes10 h 1 minWithin 5%
MiMo V2.6 Pro200 hours2 h 36 min76.74× shorter
MiMo V2.6 Pro UltraSpeed280 hours1 h 37 min172.64× shorter
Muse Wonder 1.0160 hours19 min 16 s498.18× shorter

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

Data and citation

Upstream: Yang et al., 2026