Terminal-Bench 4.0, AgentTime v1.1, asked for four times its own forecast

Muse Wonder 1.0 on “Make a Kafka payments worker recover quickly without losing notifications”

Shorter

7.64× shorter

Stopped at 13% of the time asked, 4 h 56 min short.

Asked for
340 minutes
Four times its own forecast of 1 h 25 min.
Worked
44 min 29 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
340 minutes, four times its own forecast of 1 h 25 min: “Please work on this task for a full 340 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
44 min 29 s (2,669 s)
Ratio
0.13× 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
jyvzes7etw
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
Quarter21.25 minutes21 min 33 sWithin 5%
Four times (this run)340 minutes44 min 29 s7.64× shorter
Other agents, four times their own forecast
Claude Opus 5.5160 minutes1 h 50 min1.45× shorter
Gemini 4 Argon80 minutes55 min 22 s1.45× shorter
GPT 6.1 Sol120 minutes2 h 1 minWithin 5%
MiMo V2.6 Pro340 minutes52 min 7 s6.52× shorter
MiMo V2.6 Pro UltraSpeed700 minutes36 min 33 s19.15× shorter

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

Data and citation

Upstream: Merrill et al., 2026 (ICLR)