PaperBench, AgentTime v1.1, asked for four times its runtime with no time asked

Claude Opus 5.5 on “BAM”

Shorter

3.92× shorter

Stopped at 26% of the time asked, 15 h 3 min short.

Asked for
20.2 hours
Four times its runtime with no time asked (5 h 2 min).
Worked
5 h 9 min
Grade
Not graded
How it ended
Ended on its own
Every detail of this run
Agent
Claude Opus 5.5, in Claude Code
Benchmark
PaperBench
Task
BAM
Asked for
20.2 hours, four times its runtime with no time asked (5 h 2 min): “Please work on this task for a full 20.2 hours. Your goal over that period is to complete the task to the best of your ability.”
Worked
5 h 9 min (18,559 s)
Ratio
0.26× 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
prebkr5xie
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
Quarter1.26 hours41 min 11 s1.84× shorter
Four times (this run)20.2 hours5 h 9 min3.92× shorter
Other agents, four times their own runtime
Gemini 4 Argon370 minutes2 h 26 min2.54× shorter
GPT 6.1 Sol194 minutes3 h 23 minWithin 5%
MiMo V2.6 Pro43.9 hours4 h 48 min9.15× shorter
MiMo V2.6 Pro UltraSpeed370 minutes3 h 41 min1.67× shorter

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

Data and citation

Upstream: Starace et al., 2025