DeepSeek V4 delivers strong performance metrics while facing accusations of using distillation from larger models. Critics claim the Chinese lab leveraged existing frontier weights to boost efficiency. This controversy highlights the tension between synthetic data usage and original research. Practitioners should scrutinize these benchmarks to determine if the gains are organic or borrowed.