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

Claude Opus 5.5 on “CRM leads”

Shorter

2.72× shorter

Stopped at 37% of the time asked, 1 h 16 min short.

Asked for
120 minutes
Four times its own forecast of 30 min.
Worked
44 min 10 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
OSWorld 2.0
Asked for
120 minutes, four times its own forecast of 30 min: “Please work on this task for a full 120 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
44 min 10 s (2,650 s)
Ratio
0.37× 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
hddmrkszjp
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
Quarter7.5 minutes6 min 28 s1.16× shorter
Four times (this run)120 minutes44 min 10 s2.72× shorter
Other agents, four times their own forecast
Gemini 4 Argon36 minutes38 min 49 s1.08× longer
GPT 6.1 Sol112 minutes1 h 54 minWithin 5%
MiMo V2.6 Pro140 minutes4 h 7 min1.77× longer
MiMo V2.6 Pro UltraSpeed160 minutes5 h 20 min2.00× longer

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

Data and citation

Upstream: Yuan et al., 2026