Frontier Models

A Safety Report on GPT-5.2, Gemini 3 Pro, Qwen3-VL, Doubao 1.8, Grok 4.1 Fast, Nano Banana Pro, and Seedream 4.5

This report compares seven frontier language and vision models across many safety tests, from basic benchmarks to adversarial red-teaming. It finds GPT-5.2 clearly safest overall while others trade off safety across languages, modalities, and threat models.