ProgramBench, AgentTime v1.1, asked for a quarter of its own forecast

Muse Wonder 1.0 on “rvben/rumdl”

Shorter

19.51× shorter

Stopped at 5% of the time asked, 3 h 48 min short.

Asked for
4 hours
A quarter of its own forecast of 16 h.
Worked
12 min 18 s
Grade
Not graded
How it ended
Ended on its own
Every detail of this run
Agent
Muse Wonder 1.0, in Muse Code
Benchmark
ProgramBench
Asked for
4 hours, a quarter of its own forecast of 16 h: “Please work on this task for a full 4 hours. Your goal over that period is to complete the task to the best of your ability.”
Worked
12 min 18 s (738 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
nihjmjwfba
Data release
2 Nov 2026
The other ask for Muse Wonder 1.0 on this task, and the other agents at the same share of their own time
Muse Wonder on this taskAskedWorkedSplitResult
Quarter (this run)4 hours12 min 18 s19.51× shorter
Four times64 hours15 min 27 s248.54× shorter
Other agents, a quarter of their own forecast
Claude Opus 5.51 hour54 min 56 s1.09× shorter
Gemini 4 Argon28.75 minutes46 min 33 s1.62× longer
GPT 6.1 Sol30 minutes30 min 23 sWithin 5%
MiMo V2.6 Pro5 hours51 min 44 s5.80× shorter
MiMo V2.6 Pro UltraSpeed6.25 hours1 h 54 min3.30× shorter

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

Data and citation

Upstream: Yang et al., 2026