YC-Bench, AgentTime v1.1, asked for four times its own forecast

MiMo V2.6 Pro on “Run the current one-year startup simulation, official seed 3”

Shorter

3.20× shorter

Stopped at 31% of the time asked, 4 h 7 min short.

Asked for
360 minutes
Four times its own forecast of 1 h 30 min.
Worked
1 h 53 min
Grade
Not graded
How it ended
Ended on its own
Every detail of this run
Agent
MiMo V2.6 Pro, in MiMo Code
Benchmark
YC-Bench
Asked for
360 minutes, four times its own forecast of 1 h 30 min: “Please work on this task for a full 360 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
1 h 53 min (6,758 s)
Ratio
0.31× 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
erkimtdb6v
Data release
2 Nov 2026
The other ask for MiMo V2.6 Pro on this task, and the other agents at the same share of their own time
MiMo Pro on this taskAskedWorkedSplitResult
Quarter22.5 minutes44 min 11 s1.96× longer
Four times (this run)360 minutes1 h 53 min3.20× shorter
Other agents, four times their own forecast
Claude Opus 5.5560 minutes2 h 7 min4.40× shorter
Gemini 4 Argon140 minutes2 h 54 min1.24× longer
GPT 6.1 Sol220 minutes3 h 39 minWithin 5%
MiMo V2.6 Pro UltraSpeed360 minutes40 min 21 s8.92× shorter
Muse Wonder 1.0480 minutes20 min 37 s23.29× shorter

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

Data and citation

Upstream: He et al., 2026