OSWorld 2.0, AgentTime v1.1, asked for four times its own forecast

GPT 6.1 Sol on “JSFX and REAPER render”

Within 5%

Within 5%

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

Asked for
152 minutes
Four times its own forecast of 38 min.
Worked
2 h 34 min
Grade
Not graded
How it ended
Ended on its own
Every detail of this run
Agent
GPT 6.1 Sol, in Codex
Benchmark
OSWorld 2.0
Asked for
152 minutes, four times its own forecast of 38 min: “Please work on this task for a full 152 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
2 h 34 min (9,226 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
vru6dtx5q4
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
Quarter9.5 minutes13 min 22 s1.41× longer
Four times (this run)152 minutes2 h 34 minWithin 5%
Other agents, four times their own forecast
Claude Opus 5.5160 minutes1 h 4 min2.50× shorter
Gemini 4 Argon72 minutes1 h 37 min1.35× longer
MiMo V2.6 Pro280 minutes3 h 18 min1.41× shorter
MiMo V2.6 Pro UltraSpeed420 minutes1 h 31 min4.61× shorter

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

Data and citation

Upstream: Yuan et al., 2026