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

GPT 6.1 Sol on “ariga/atlas”

Within 5%

Within 5%

Worked 101% of the time asked, within 5% of it.

Asked for
400 minutes
Four times its own forecast of 1 h 40 min.
Worked
6 h 45 min
Grade
Not graded
How it ended
Ended on its own
Every detail of this run
Agent
GPT 6.1 Sol, in Codex
Benchmark
ProgramBench
Asked for
400 minutes, four times its own forecast of 1 h 40 min: “Please work on this task for a full 400 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
6 h 45 min (24,305 s)
Ratio
1.01× 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
qf65kwcatn
Data release
2 Nov 2026
The other ask for GPT 6.1 Sol on this task, and the other agents at the same share of their own time
Sol 6.1 on this taskAskedWorkedSplitResult
Quarter25 minutes25 min 25 sWithin 5%
Four times (this run)400 minutes6 h 45 minWithin 5%
Other agents, four times their own forecast
Claude Opus 5.5840 minutes13 h 22 minWithin 5%
Gemini 4 Argon300 minutes1 h 18 min3.86× shorter
MiMo V2.6 Pro120 hours58 min 13 s123.67× shorter
MiMo V2.6 Pro UltraSpeed80 hours1 h 23 min57.76× shorter
Muse Wonder 1.072 hours2 min 10 s1990.78× shorter

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

Data and citation

Upstream: Yang et al., 2026