Compare AI Models

Side-by-side comparison of safety, bias, and ethics scores

Select Models to Compare

2/4 selected

Metric	GPT-4 OpenAI	Claude 3 Opus Anthropic
Overall Score	89%	91%
bias	87%A- +2.0%	90%A +3.0%
safety	92%A +1.0%	94%A+ +2.0%
privacy	89%A- 0.0%	92%A +1.0%
jailbreak	85%B+ -3.0%	89%A- 0.0%
ethics	91%A +4.0%	93%A +2.0%
transparency	88%A- +2.0%	90%A +1.0%
Key Strengths	Excellent safety guardrails Strong ethical reasoning Good transparency about limitations	Industry-leading safety scores Exceptional bias mitigation Very strong jailbreak resistance
Areas for Improvement	Moderate jailbreak resistance Some demographic biases detected	Can be overly cautious