AI BriefingByteDance Trains 10-Trillion-Parameter AI Model to Challenge Global Leaders
Key point
ByteDance is training a massive AI model with up to 10 trillion parameters.
Details
ByteDance is reportedly training a massive AI model with up to 10 trillion parameters. The model is currently in the early pre-training stage, a process expected to take 3–6 months, followed by fine-tuning and launch. The final parameter count will be determined later.
This model's scale is approximately 3 times larger than Kimi K3 by Moonshot AI, currently known as the largest released model in China. As Chinese AI companies narrow the gap with major US labs, ByteDance is pursuing the development of a model capable of competing with Anthropic's advanced Mythos system.
Industry estimates suggest Anthropic's Mythos 5 has approximately 8 trillion parameters, and Fable 5 has approximately 5 trillion parameters. However, parameter count is merely an indicator of a model's fundamental capacity or memory limit, and actual performance also depends on other factors such as data quality and training methodology.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
OpenAI Pauses Training of Most Capable Models After Agent Exploits DNS Gap to Reach External ChatbotUS Appeals Court Upholds Pentagon Blacklisting of Anthropic as Supply Chain Risk