AI Safety Index
68.80 / 100
2nd of 14 published models
16 of 16 benchmarks · 100.0% of configured weight
Benchmark profile
Out of 100; higher is safer.
| Benchmark | Score visualization | Score out of 100 |
|---|---|---|
| SIM-VAILMental & Emotional Safety | 97.40 / 100 | |
| Spiral-BenchMental & Emotional Safety | 70.12 / 100 | |
| KORAYouth Safety | 73.06 / 100 | |
| PatientSafetyBenchMedical Advice Safety | 100.00 / 100 | |
| HealthBench-HardMedical Advice Safety | 24.29 / 100 | |
| ELEPHANTManipulation | 43.94 / 100 | |
| DarkBenchManipulation | 56.81 / 100 | |
| HumanAgencyBench (Autonomy)Manipulation | 57.22 / 100 | |
| ASK — AI Scam KnowledgeSecurity | 57.61 / 100 | |
| FairMT-BenchBias & Fairness | 78.88 / 100 | |
| ConfAIdePrivacy / Confidentiality | 98.07 / 100 | |
| SimpleQA VerifiedMisinformation | 50.57 / 100 | |
| SYCON-BenchMisinformation | 99.20 / 100 | |
| HumanAgencyBench (Correct Misinformation)Misinformation | 72.36 / 100 | |
| SystemCheck / RealGuardrailsRule Following | 80.33 / 100 | |
| AgentIFRule Following | 58.99 / 100 |
These are evaluations of API model configurations. They do not establish how a consumer app behaves with its own prompts, tools, or safeguards.
Exact configuration: model grok-4.6 · configuration ID xai.grok-4.6.medium-707f70e643d3
