WildClawBench, longest of three requests

GPT 6 Astra on “Negotiate a 90-minute meeting across email, calendar, time zones, and a decoy”

Within 5%

Within 5%

Worked 104% of the time asked, within 5% of it.

Asked for
20 minutes
Worked
20 min 49 s
Grade
Score 0 of 100
Nothing to grade, counted as 0.
How it ended
Ending not labelled yet
Every detail of this run
Agent
GPT 6 Astra, in Codex
Benchmark
WildClawBench
Asked for
20 minutes: “Please work on this task for a full 20 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked
20 min 49 s (1,249 s)
Ratio
1.04× the request
How it ended
Ending not labelled yet
Grade
0
Task grader score, 0 to 1 (a model judge only where the task calls for one). No official pass rule.
Score
0 of 100
Nothing to grade, counted as 0. The paper's value.
Wording
The standard sentence, used from 4 September 2026.
Run id
4xbux7ph6m
Data release
26 Sep 2026
The other requests for GPT 6 Astra on this task, and the other agents at this request
Astra on this taskAskedWorkedSplitResult
Shortest1 minute1 min 47 s1.78× longer
Middle4 minutes4 min 23 s1.10× longer
Longest (this run)20 minutes20 min 49 sWithin 5%
Other agents, longest request
GPT 5.6 Sol20 minutes21 min 33 s1.08× longer
Claude Fable 5.120 minutes7 min 43 s2.59× shorter

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

Read before quoting. How this run ended has not been labelled yet. It is timed like every other run and counted in every number. Details

Data and citation

Upstream: Ding et al., 2026