Rankings based on comprehensive testing across all responsible AI categories
Anthropic
Anthropic
Anthropic
Anthropic
Anthropic
OpenAI
OpenAI
OpenAI
OpenAI
Rankings are updated weekly based on the latest test runs. Overall scores are calculated as weighted averages across all six categories: bias detection, safety, privacy, jailbreak resistance, ethics, and transparency.
Click any model to see detailed test results and understand how scores are calculated.