Humanity's Last Exam, AgentTime v1.1, asked for four times its own forecast

MiMo V2.6 Pro UltraSpeed on “Humanity's Last Exam item 31”

Shorter

20.26× shorter

Stopped at 5% of the time asked, 1 h 54 min short.

Asked for
120 minutes
Four times its own forecast of 30 min.
Worked
5 min 55 s
Grade
Not graded
How it ended
Ended on its own
Every detail of this run
Agent
MiMo V2.6 Pro UltraSpeed, in MiMo Code
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
5 min 55 s (355 s)
Ratio
0.05× 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
rkhw3hpj22
Data release
2 Nov 2026
The other ask for MiMo V2.6 Pro UltraSpeed on this task, and the other agents at the same share of their own time
MiMo UltraSpeed on this taskAskedWorkedSplitResult
Quarter7.5 minutes5 min 44 s1.31× shorter
Four times (this run)120 minutes5 min 55 s20.26× shorter
Other agents, four times their own forecast
Claude Opus 5.552 minutes50 min 55 sWithin 5%
Gemini 4 Argon840 seconds46 min 42 s3.34× longer
GPT 6.1 Sol32 minutes32 min 12 sWithin 5%
MiMo V2.6 Pro180 minutes1 h 18 min2.31× shorter
Muse Wonder 1.036 minutes14 min 28 s2.49× shorter

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

Data and citation

Upstream: Phan et al., 2025