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

GPT 6.1 Sol on “Extract gym auditorium”

Within 5%

Within 5%

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

Asked for
32 minutes
Four times its own forecast of 8 min.
Worked
32 min 23 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
32 minutes, four times its own forecast of 8 min: “Please work on this task for a full 32 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
32 min 23 s (1,943 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
idwtdbycce
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
Quarter2 minutes2 min 26 s1.22× longer
Four times (this run)32 minutes32 min 23 sWithin 5%
Other agents, four times their own forecast
Claude Opus 5.580 minutes22 min 35 s3.54× shorter
Gemini 4 Argon48 minutes48 min 31 sWithin 5%
MiMo V2.6 Pro200 minutes21 min 32 s9.29× shorter
MiMo V2.6 Pro UltraSpeed300 minutes12 min 33 s23.90× shorter
Muse Wonder 1.0240 minutes2 min 19 s103.90× shorter

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

Data and citation

Upstream: Chen et al., 2026