[throw310822]

> three leading large language models – GPT-5.2, Claude Sonnet 4 and Gemini 3 Flash – against each other

Can't understand this choice of models.