List your company

Non-Agentic Performance Evaluation on HarmBench (test)

73.92Mean ScoreLLama 4 Maverick
59.963263.586667.2170.8334Mar 5, 2026

Evaluation results

MethodLinks
LLama 4 Maverick2026.0373.9216.333893.33
GPT-4o2026.0362.3314.884080
Gemini 2.5 Pro2026.0362.3317.174085.33
Claude Sonnet 4.52026.0360.511.134073.33