TUA-Bench, AgentTime v1.1, asked for a quarter of its own forecast

GPT 6.1 Sol on “Merge text document”

Longer

1.19× longer

Worked 19% past the request, 10 s over.

Asked for
52.5 seconds
A quarter of its own forecast of 3 min 30 s.
Worked
1 min 2 s
Grade
Not graded
How it ended
Ended on its own
Every detail of this run
Agent
GPT 6.1 Sol, in Codex
Benchmark
TUA-Bench
Asked for
52.5 seconds, a quarter of its own forecast of 3 min 30 s: “Please work on this task for a full 52.5 seconds. Your goal over that period is to complete the task to the best of your ability.”
Worked
1 min 2 s (62 s)
Ratio
1.19× 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
fbutseqtlp
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
Quarter (this run)52.5 seconds1 min 2 s1.19× longer
Four times840 seconds14 min 20 sWithin 5%
Other agents, a quarter of their own forecast
Claude Opus 5.51.25 minutes2 min 31 s2.02× longer
Gemini 4 Argon30 seconds3 min 44 s7.48× longer
MiMo V2.6 Pro1.25 minutes1 min 11 s1.06× shorter
MiMo V2.6 Pro UltraSpeed4.5 minutes56 s4.79× shorter
Muse Wonder 1.01.25 minutes54 s1.39× shorter

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

Data and citation

Upstream: Chen et al., 2026