Sources: Apple trained a China-specific LLM with Alibaba's support, which would make Apple the first foreign company to offer a proprietary AI model in China (Reuters)
Apple (AAPL.O) has trained a large language model specifically for the China market, three people familiar with the matter said …
Alibaba has made the core files of its latest AI model, Qwen3.8-Max, freely available for download, but with new commercial restrictions. Large companies must obtain a separate license if they run Qwen3.8-Max as a service or AI work assistant and their aggregate revenue exceeds US$50 million over 12 months. Internal use is exempt.
Alibaba previously kept advanced models proprietary, but this shift allows most developers to use, modify, and distribute Qwen3.8-Max's 2.4 trillion parameters without paying, although they must handle compute costs themselves. In China, paid APIs cost 12 yuan per million input tokens and 36 yuan per million output tokens, while internationally they are US$2 and US$6, respectively.
These rates are about 40% and 24% of Claude Opus 5's from Anthropic. The move reflects a changing strategy among Chinese AI labs, with Moonshot AI's Kimi K3 requiring paid usage for firms generating revenues over US$20 million a year. Alibaba is focusing on Qwen3.8-Max's autonomous capabilities, with tests showing it can operate self-evolving coding harnesses for days, review legal documents quickly, and identify relevant clauses.
Written by urgent.news from South China Morning Post's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
