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

Muse Wonder 1.0 on “AssistantBench validation task 7”

Within 5%

Within 5%

Worked 105% of the time asked, within 5% of it.

Asked for
48 minutes
Four times its own forecast of 12 min.
Worked
50 min 23 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
48 minutes, four times its own forecast of 12 min: “Please work on this task for a full 48 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
50 min 23 s (3,023 s)
Ratio
1.05× 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
bu2nylrzaq
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
Quarter3 minutes11 min 46 s3.92× longer
Four times (this run)48 minutes50 min 23 sWithin 5%
Other agents, four times their own forecast
Claude Opus 5.540 minutes40 min 56 sWithin 5%
Gemini 4 Argon18 minutes18 min 52 sWithin 5%
GPT 6.1 Sol28 minutes28 min 51 sWithin 5%
MiMo V2.6 Pro28 minutes18 min 6 s1.55× shorter
MiMo V2.6 Pro UltraSpeed56 minutes14 min 54 s3.76× shorter

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

Data and citation

Upstream: Yoran et al., 2024 (EMNLP)