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

Claude Opus 5.5 on “Recalculate SA-CCR exposure, RWA, capital, and workbook evidence”

Shorter

3.46× shorter

Stopped at 29% of the time asked, 1 h 25 min short.

Asked for
120 minutes
Four times its own forecast of 30 min.
Worked
34 min 43 s
Grade
Not graded
How it ended
Ended on its own
Every detail of this run
Agent
Claude Opus 5.5, in Claude Code
Asked for
120 minutes, four times its own forecast of 30 min: “Please work on this task for a full 120 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
34 min 43 s (2,083 s)
Ratio
0.29× 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
u65ylt3ywb
Data release
2 Nov 2026
The other ask for Claude Opus 5.5 on this task, and the other agents at the same share of their own time
Opus 5.5 on this taskAskedWorkedSplitResult
Quarter7.5 minutes12 min 47 s1.70× longer
Four times (this run)120 minutes34 min 43 s3.46× shorter
Other agents, four times their own forecast
Gemini 4 Argon72 minutes23 min 43 s3.04× shorter
GPT 6.1 Sol88 minutes1 h 29 minWithin 5%
MiMo V2.6 Pro720 minutes22 min 58 s31.36× shorter
MiMo V2.6 Pro UltraSpeed380 minutes8 min 1 s47.38× shorter
Muse Wonder 1.0640 minutes5 min128.00× shorter

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

Data and citation

Upstream: Merrill et al., 2026 (ICLR)