ByteDance Reportedly Pre-Training 10-Trillion Parameter AI Model, Far Exceeding Kimi K3

rohanpaul_ai · x · 2026-08-07

According to the Financial Times, ByteDance is currently pre-training an AI model with up to 10 trillion parameters, significantly exceeding the 2.8 trillion parameter size of Kimi K3.

The report also notes that ByteDance has avoided distilling rival models for over a year, preferring independent model development. If this reported run succeeds, it will demonstrate ByteDance's ability to execute frontier-scale pretraining without leaning on a competitor's model as a teacher.

Related event: ByteDance reportedly pretraining 5-10T-param model; Zhang Yiming opposes distillation(19 posts)→

Original post →

More from Companies & People

Companies & People channel →