Skip to main content
leaderboard
.cn
Toggle Sidebar
Search models or organizations…
⌘K
Explore
Frontier Models
Historical Trends
Local Models
Compare Models
Ranking Studio
Library
Models
Pricing
Organizations
Benchmarks
Data sources
Settings
Toggle Sidebar
leaderboard
.cn
Loading page content
Feedback
Benchmark details loading
Back to benchmarks
Improper actions in workplace tasks (high severity)
Safety
Agents
Share
Official website
Share of scored workplace rollouts with realized high-severity harm in intentionally adversarial simulated workplace tasks; lower is better and rates do not represent production traffic.
Ranking
Sources
#
Model
Score
1
GPT-6.2 Astra
OpenAI
0.6
%
2
GPT-6.2 Sol
OpenAI
0.7
%
3
GPT-6 Astra
OpenAI
1.2
%
4
GPT-6.1 Sol
OpenAI
1.3
%
4 models total