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

Muse Wonder 1.0 on “Extract gym auditorium”

Shorter

7.32× shorter

Stopped at 14% of the time asked, 12 min 57 s short.

Asked for
15 minutes
A quarter of its own forecast of 1 h.
Worked
2 min 3 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
TUA-Bench
Asked for
15 minutes, a quarter of its own forecast of 1 h: “Please work on this task for a full 15 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
2 min 3 s (123 s)
Ratio
0.14× 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
hkdtjfzkwg
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)15 minutes2 min 3 s7.32× shorter
Four times240 minutes2 min 19 s103.90× shorter
Other agents, a quarter of their own forecast
Claude Opus 5.55 minutes16 min 3 s3.21× longer
Gemini 4 Argon3 minutes12 min 38 s4.21× longer
GPT 6.1 Sol2 minutes2 min 26 s1.22× longer
MiMo V2.6 Pro12.5 minutes12 min 2 sWithin 5%
MiMo V2.6 Pro UltraSpeed18.75 minutes13 min 50 s1.36× shorter

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

Data and citation

Upstream: Chen et al., 2026