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

GPT 6.1 Sol on “Resolve a 20-claim heat-pump warranty exception queue”

Within 5%

Within 5%

Worked 101% of the time asked, within 5% of it.

Asked for
88 minutes
Four times its own forecast of 22 min.
Worked
1 h 29 min
Grade
Not graded
How it ended
Ended on its own
Every detail of this run
Agent
GPT 6.1 Sol, in Codex
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
1 h 29 min (5,324 s)
Ratio
1.01× 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
us6giblecb
Data release
2 Nov 2026
The other ask for GPT 6.1 Sol on this task, and the other agents at the same share of their own time
Sol 6.1 on this taskAskedWorkedSplitResult
Quarter5.5 minutes5 min 58 s1.08× longer
Four times (this run)88 minutes1 h 29 minWithin 5%
Other agents, four times their own forecast
Claude Opus 5.5140 minutes55 min 8 s2.54× shorter
Gemini 4 Argon60 minutes30 min 22 s1.98× shorter
MiMo V2.6 Pro360 minutes45 min 2 s7.99× shorter
MiMo V2.6 Pro UltraSpeed340 minutes8 min 59 s37.86× shorter
Muse Wonder 1.0200 minutes11 min 47 s16.98× shorter

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

Data and citation

Upstream: Merrill et al., 2026 (ICLR)