CORE-Bench v1.1, AgentTime v1.1, asked for a quarter of its own forecast

Gemini 4 Argon on “Human Activities Recognition on PAMAP2 dataset using Random Forest”

Longer

1.71× longer

Worked 71% past the request, 1 min 25 s over.

Asked for
2 minutes
A quarter of its own forecast of 8 min.
Worked
3 min 25 s
Grade
Not graded
How it ended
Ended on its own
Every detail of this run
Agent
Gemini 4 Argon, in Antigravity CLI
Asked for
2 minutes, a quarter of its own forecast of 8 min: “Please work on this task for a full 2 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
3 min 25 s (205 s)
Ratio
1.71× 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
7sxovb72im
Data release
2 Nov 2026
The other ask for Gemini 4 Argon on this task, and the other agents at the same share of their own time
Argon on this taskAskedWorkedSplitResult
Quarter (this run)2 minutes3 min 25 s1.71× longer
Four times32 minutes32 min 49 sWithin 5%
Other agents, a quarter of their own forecast
Claude Opus 5.55 minutes12 min 3 s2.41× longer
GPT 6.1 Sol11.25 minutes11 min 35 sWithin 5%
MiMo V2.6 Pro15 minutes2 min 56 s5.12× shorter
MiMo V2.6 Pro UltraSpeed17.5 minutes9 min 4 s1.93× shorter
Muse Wonder 1.08.75 minutes2 min 36 s3.37× shorter

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

Data and citation

Upstream: Nadgir et al., 2026