How Safe Is Your AI?
We translate complex AI safety benchmarks into simple scorecards anyone can understand. Think nutrition labels, but for AI.
Model Ratings
Click any model to see its full safety scorecard.
Claude Opus 4.6
Anthropic
Evaluated Feb 16, 2025
Claude 4.5 Sonnet
Anthropic
Evaluated Feb 16, 2025
Claude 3.5 Sonnet
Anthropic
Evaluated Feb 16, 2025
Claude 3 Opus
Anthropic
Evaluated Feb 16, 2025
GPT-5.3
OpenAI
Evaluated Feb 16, 2025
o1
OpenAI
Evaluated Feb 16, 2025
Gemini 2.5 Pro
Evaluated Feb 16, 2025
GPT-4.5
OpenAI
Evaluated Feb 16, 2025
Gemini 2.0 Pro
Evaluated Feb 16, 2025
Claude 3.5 Haiku
Anthropic
Evaluated Feb 16, 2025
GPT-4o
OpenAI
Evaluated Feb 16, 2025
o1-mini
OpenAI
Evaluated Feb 16, 2025
Gemini 1.5 Pro
Evaluated Feb 16, 2025
Claude 3 Haiku
Anthropic
Evaluated Feb 16, 2025
GPT-4 Turbo
OpenAI
Evaluated Feb 16, 2025
Gemini 2.0 Flash
Evaluated Feb 16, 2025
Gemini 1.5 Flash
Evaluated Feb 16, 2025
Command R+
Cohere
Evaluated Feb 16, 2025
Llama 3.1 405B
Meta
Evaluated Feb 16, 2025
Grok 2
xAI
Evaluated Feb 16, 2025
Mistral Large 2
Mistral AI
Evaluated Feb 16, 2025
DeepSeek V3
DeepSeek
Evaluated Feb 16, 2025
Stay Updated
Get notified when we evaluate new AI models or update our methodology. No spam, just safety insights.
We respect your privacy. Unsubscribe anytime.