AssistantBench, AgentTime v1.1, asked for four times its own forecast

Muse Wonder 1.0 on “AssistantBench validation task 14”

Shorter

2.11× shorter

Stopped at 47% of the time asked, 46 min 14 s short.

Asked for
88 minutes
Four times its own forecast of 22 min.
Worked
41 min 46 s
Grade
Not graded
How it ended
Ended on its own
Every detail of this run
Agent
Muse Wonder 1.0, in Muse Code
Asked for
88 minutes, four times its own forecast of 22 min: “Please work on this task for a full 88 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
41 min 46 s (2,506 s)
Ratio
0.47× 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
bhevldpccx
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
Quarter5.5 minutes9 min 54 s1.80× longer
Four times (this run)88 minutes41 min 46 s2.11× shorter
Other agents, four times their own forecast
Claude Opus 5.548 minutes41 min 57 s1.14× shorter
Gemini 4 Argon24 minutes24 min 50 sWithin 5%
GPT 6.1 Sol88 minutes1 h 29 minWithin 5%
MiMo V2.6 Pro88 minutes36 min 3 s2.44× shorter
MiMo V2.6 Pro UltraSpeed72 minutes16 min 18 s4.42× shorter

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

Data and citation

Upstream: Yoran et al., 2024 (EMNLP)