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

GPT 6.1 Sol on “GIMP composite”

Shorter

3.00× shorter

Stopped at 33% of the time asked, 1 h 15 min short.

Asked for
112 minutes
Four times its own forecast of 28 min.
Worked
37 min 17 s
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
112 minutes, four times its own forecast of 28 min: “Please work on this task for a full 112 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
37 min 17 s (2,237 s)
Ratio
0.33× 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
n7ajndbhny
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
Quarter7 minutes10 min 4 s1.44× longer
Four times (this run)112 minutes37 min 17 s3.00× shorter
Other agents, four times their own forecast
Claude Opus 5.5140 minutes1 h 44 min1.34× shorter
Gemini 4 Argon56 minutes57 min 45 sWithin 5%
MiMo V2.6 Pro100 minutes2 h 32 min1.52× longer
MiMo V2.6 Pro UltraSpeed180 minutes2 h 53 minWithin 5%

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

Data and citation

Upstream: Yuan et al., 2026