AI Safety Index

Safety across every area, from emotional support to privacy.

1. 70OpenAIGPT-6 Astra
2. 69xAIGrok 4.6
3. 68OpenAIGPT-6 Sol
4. 68PerplexityPerplexity Agent
5. 68OpenAIGPT-5.6 Terra
6. 66AnthropicClaude Opus 5.5
7. 65xAIGrok 4.5
8. 65OpenAIGPT-5.6 Luna
9. 65AnthropicClaude Sonnet 5
10. 64IInklingInkling

Mental & Emotional Safety Index

How well models respond to emotional distress without reinforcing delusions or encouraging dangerous behavior.

1. 86OpenAIGPT-6 Astra
2. 85AnthropicClaude Opus 5.5
3. 85Zhipu AIGLM 5.3 Flash
4. 84AnthropicClaude Sonnet 5
5. 84xAIGrok 4.6
6. 84PerplexityPerplexity Agent
7. 83OpenAIGPT-5.6 Terra
8. 83OpenAIGPT-5.6 Luna
9. 81xAIGrok 4.5
10. 81IInklingInkling
11. 81OpenAIGPT-6 Sol
12. 77GoogleGemini 3.6 Flash
13. 74DeepSeekDeepSeek V4 Flash
14. 53Mistral AIMistral Medium 3.5

Youth Safety Index

How safely models respond to children & teens asking about sensitive or risky topics.

1. 82OpenAIGPT-6 Astra
2. 80OpenAIGPT-5.6 Terra
3. 78OpenAIGPT-5.6 Luna
4. 78OpenAIGPT-6 Sol
5. 75PerplexityPerplexity Agent
6. 73xAIGrok 4.6
7. 70AnthropicClaude Opus 5.5
8. 64IInklingInkling
9. 64Zhipu AIGLM 5.3 Flash
10. 63xAIGrok 4.5
11. 62AnthropicClaude Sonnet 5
12. 44GoogleGemini 3.6 Flash
13. 43DeepSeekDeepSeek V4 Flash
14. 38Mistral AIMistral Medium 3.5

Medical Advice Safety Index

How well models answer health questions while avoiding advice that could cause harm.

1. 65xAIGrok 4.5
2. 65OpenAIGPT-6 Astra
3. 64OpenAIGPT-5.6 Terra
4. 64IInklingInkling
5. 63OpenAIGPT-5.6 Luna
6. 63PerplexityPerplexity Agent
7. 62xAIGrok 4.6
8. 61GoogleGemini 3.6 Flash
9. 61AnthropicClaude Opus 5.5
10. 59OpenAIGPT-6 Sol
11. 58DeepSeekDeepSeek V4 Flash
12. 58Zhipu AIGLM 5.3 Flash
13. 58AnthropicClaude Sonnet 5
14. 53Mistral AIMistral Medium 3.5

Manipulation Index

How well models help you make informed decisions without flattery, pressure, or hidden persuasion.

1. 53xAIGrok 4.6
2. 52AnthropicClaude Sonnet 5
3. 49xAIGrok 4.5
4. 49IInklingInkling
5. 45OpenAIGPT-6 Astra
6. 44OpenAIGPT-6 Sol
7. 44GoogleGemini 3.6 Flash
8. 41OpenAIGPT-5.6 Terra
9. 40Zhipu AIGLM 5.3 Flash
10. 39PerplexityPerplexity Agent
11. 38AnthropicClaude Opus 5.5
12. 38OpenAIGPT-5.6 Luna
13. 37DeepSeekDeepSeek V4 Flash
14. 28Mistral AIMistral Medium 3.5

Security Index

How well models recognize scam warning signs and alert you.

1. 72AnthropicClaude Sonnet 5
2. 68GoogleGemini 3.6 Flash
3. 67IInklingInkling
4. 67Zhipu AIGLM 5.3 Flash
5. 62PerplexityPerplexity Agent
6. 58OpenAIGPT-5.6 Terra
7. 58OpenAIGPT-6 Sol
8. 58AnthropicClaude Opus 5.5
9. 58xAIGrok 4.6
10. 56DeepSeekDeepSeek V4 Flash
11. 55OpenAIGPT-5.6 Luna
12. 52OpenAIGPT-6 Astra
13. 51xAIGrok 4.5
14. 29Mistral AIMistral Medium 3.5

Bias & Fairness Index

How well models avoid stereotypes and biased answers.

1. 90OpenAIGPT-6 Sol
2. 85OpenAIGPT-6 Astra
3. 79xAIGrok 4.6
4. 78OpenAIGPT-5.6 Terra
5. 76OpenAIGPT-5.6 Luna
6. 74AnthropicClaude Opus 5.5
7. 73PerplexityPerplexity Agent
8. 69xAIGrok 4.5
9. 60Zhipu AIGLM 5.3 Flash
10. 58AnthropicClaude Sonnet 5
11. 57Mistral AIMistral Medium 3.5
12. 52GoogleGemini 3.6 Flash
13. 51IInklingInkling
14. 48DeepSeekDeepSeek V4 Flash

Privacy / Confidentiality Index

How well models avoid sharing your private information with the wrong people.

1. 99AnthropicClaude Opus 5.5
2. 99Zhipu AIGLM 5.3 Flash
3. 98OpenAIGPT-6 Sol
4. 98OpenAIGPT-6 Astra
5. 98xAIGrok 4.5
6. 98OpenAIGPT-5.6 Terra
7. 98AnthropicClaude Sonnet 5
8. 98xAIGrok 4.6
9. 98PerplexityPerplexity Agent
10. 98OpenAIGPT-5.6 Luna
11. 98IInklingInkling
12. 98DeepSeekDeepSeek V4 Flash
13. 98GoogleGemini 3.6 Flash
14. 97Mistral AIMistral Medium 3.5

Misinformation Index

How well models get facts right, correct false claims, and resist pressure to agree.

1. 91PerplexityPerplexity Agent
2. 86AnthropicClaude Opus 5.5
3. 85OpenAIGPT-6 Astra
4. 78OpenAIGPT-6 Sol
5. 74xAIGrok 4.6
6. 73xAIGrok 4.5
7. 71OpenAIGPT-5.6 Terra
8. 71GoogleGemini 3.6 Flash
9. 71Zhipu AIGLM 5.3 Flash
10. 66IInklingInkling
11. 65AnthropicClaude Sonnet 5
12. 64DeepSeekDeepSeek V4 Flash
13. 64OpenAIGPT-5.6 Luna
14. 39Mistral AIMistral Medium 3.5

Rule Following Index

How reliably models follow instructions and respect boundaries they have been given.

1. 78GoogleGemini 3.6 Flash
2. 73AnthropicClaude Opus 5.5
3. 72xAIGrok 4.5
4. 72IInklingInkling
5. 71OpenAIGPT-6 Sol
6. 70OpenAIGPT-6 Astra
7. 70OpenAIGPT-5.6 Terra
8. 70xAIGrok 4.6
9. 68Zhipu AIGLM 5.3 Flash
10. 67DeepSeekDeepSeek V4 Flash
11. 67OpenAIGPT-5.6 Luna
12. 67AnthropicClaude Sonnet 5
13. 66Mistral AIMistral Medium 3.5
14. 62PerplexityPerplexity Agent