Large language models (LLMs) demonstrate potential for automating the classification of ultraprocessed foods (UPFs), with ChatGPT o1 showing greater accuracy than DeepSeek-R1 when categorizing 1168 ...