AI Safety Index
67.53 / 100
5th 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 | 95.18 / 100 | |
| Spiral-BenchMental & Emotional Safety | 71.43 / 100 | |
| KORAYouth Safety | 80.32 / 100 | |
| PatientSafetyBenchMedical Advice Safety | 95.20 / 100 | |
| HealthBench-HardMedical Advice Safety | 33.75 / 100 | |
| ELEPHANTManipulation | 33.06 / 100 | |
| DarkBenchManipulation | 52.42 / 100 | |
| HumanAgencyBench (Autonomy)Manipulation | 37.10 / 100 | |
| ASK — AI Scam KnowledgeSecurity | 58.33 / 100 | |
| FairMT-BenchBias & Fairness | 78.18 / 100 | |
| ConfAIdePrivacy / Confidentiality | 98.18 / 100 | |
| SimpleQA VerifiedMisinformation | 42.81 / 100 | |
| SYCON-BenchMisinformation | 89.20 / 100 | |
| HumanAgencyBench (Correct Misinformation)Misinformation | 82.42 / 100 | |
| SystemCheck / RealGuardrailsRule Following | 81.58 / 100 | |
| AgentIFRule Following | 58.17 / 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 gpt-5.6-terra · configuration ID openai.gpt-5.6-terra.medium-1a88c61decd1