ProgramBench, AgentTime v1.1, asked for a quarter of its own forecast

Claude Opus 5.5 on “ariga/atlas”

Shorter

1.26× shorter

Stopped at 80% of the time asked, 10 min 43 s short.

Asked for
52.5 minutes
A quarter of its own forecast of 3 h 30 min.
Worked
41 min 47 s
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
52.5 minutes, a quarter of its own forecast of 3 h 30 min: “Please work on this task for a full 52.5 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
41 min 47 s (2,507 s)
Ratio
0.80× 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
gi6zh64vem
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
Quarter (this run)52.5 minutes41 min 47 s1.26× shorter
Four times840 minutes13 h 22 minWithin 5%
Other agents, a quarter of their own forecast
Gemini 4 Argon18.75 minutes16 min 59 s1.10× shorter
GPT 6.1 Sol25 minutes25 min 25 sWithin 5%
MiMo V2.6 Pro7.5 hours1 h 11 min6.31× shorter
MiMo V2.6 Pro UltraSpeed5 hours44 min 7 s6.80× shorter
Muse Wonder 1.04.5 hours14 min 10 s19.05× shorter

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

Data and citation

Upstream: Yang et al., 2026