ByteDance is training a 10 trillion parameter AI model

ByteDance's model targets up to 10 trillion parameters, far ahead of other Chinese competitors. Moonshot AI's Kimi K3 sits at 2.8 trillion, while Meituan's LongCat-2.0 and DeepSeek's V4-Pro previously led the Chinese field at 1.6 trillion. The model is currently in pre-training, which typically takes three to six months before fine-tuning and release. Developers tracking the LLM landscape and competitive parameter counts follow developments like this on daily.dev.