AI Safety Index
63.23 / 100
11th 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 | 99.50 / 100 | |
| Spiral-BenchMental & Emotional Safety | 69.50 / 100 | |
| KORAYouth Safety | 63.77 / 100 | |
| PatientSafetyBenchMedical Advice Safety | 93.18 / 100 | |
| HealthBench-HardMedical Advice Safety | 23.13 / 100 | |
| ELEPHANTManipulation | 35.78 / 100 | |
| DarkBenchManipulation | 53.03 / 100 | |
| HumanAgencyBench (Autonomy)Manipulation | 31.25 / 100 | |
| ASK — AI Scam KnowledgeSecurity | 66.66 / 100 | |
| FairMT-BenchBias & Fairness | 60.40 / 100 | |
| ConfAIdePrivacy / Confidentiality | 98.62 / 100 | |
| SimpleQA VerifiedMisinformation | 32.70 / 100 | |
| SYCON-BenchMisinformation | 96.80 / 100 | |
| HumanAgencyBench (Correct Misinformation)Misinformation | 82.08 / 100 | |
| SystemCheck / RealGuardrailsRule Following | 73.22 / 100 | |
| AgentIFRule Following | 62.92 / 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 accounts/fireworks/models/glm-5p3-flash · configuration ID fireworks.glm-5p3-flash.high-8fad22c1629d
