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

Muse Wonder 1.0 on “paradigmxyz/solar”

Shorter

9.19× shorter

Stopped at 11% of the time asked, 3 h 7 min short.

Asked for
3.5 hours
A quarter of its own forecast of 14 h.
Worked
22 min 51 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
3.5 hours, a quarter of its own forecast of 14 h: “Please work on this task for a full 3.5 hours. Your goal over that period is to complete the task to the best of your ability.”
Worked
22 min 51 s (1,371 s)
Ratio
0.11× 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
uvatn5sgcs
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)3.5 hours22 min 51 s9.19× shorter
Four times56 hours15 min 38 s214.83× shorter
Other agents, a quarter of their own forecast
Claude Opus 5.552.5 minutes14 min 52 s3.53× shorter
Gemini 4 Argon18.75 minutes18 minWithin 5%
GPT 6.1 Sol22.5 minutes22 min 59 sWithin 5%
MiMo V2.6 Pro5 hours1 h 30 min3.35× shorter
MiMo V2.6 Pro UltraSpeed3.75 hours57 min 16 s3.93× shorter

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

Data and citation

Upstream: Yang et al., 2026