ByteDance Reportedly Training Massive AI Model Approaching Scale of Anthropic's Flagship System
N.R. Finch
The Financial Times reports that ByteDance is training an AI model approaching the scale of Anthropic's flagship Mythos system — the latest signal that Chinese tech giants are pushing into the same weight class as the world's leading AI labs.
What does the report actually say?
The Financial Times, citing people familiar with the matter, says ByteDance is training a very large AI model whose scale approaches Anthropic's most advanced system, Mythos.
Reuters says it could not independently verify the report.
This means → the claim currently rests on a single source and has not been confirmed by a second outlet — watch for follow-up verification.
Why is Anthropic the benchmark here?
Anthropic's Mythos is one of the largest frontier AI systems in the world, widely regarded as a ceiling-level reference point for the industry.
In plain terms = saying "approaching Mythos" is saying ByteDance is not aiming for a good-enough model — it is targeting the very top tier globally.
This reflects a shift among China's leading tech companies from "catching up" to "matching the frontier" in the large-model race.
What does this mean for the market?
If confirmed, ByteDance would join a small group of Chinese companies operating at model scales comparable to top US labs.
This means → the global large-model competition could tighten further — it is no longer an arms race among American firms alone; a Chinese player is entering the same weight class.
A key caveat: the report remains unverified, and training scale alone does not equal real-world model capability.
Content is for reference only, not financial advice.