Moonshot AI has built its profile on making large-scale AI systems visible to developers. Now, a Bloomberg account based on people familiar with the matter has put a different kind of visibility in prospect: a possible public-market examination of the Beijing company.
Xiaomi has supplied a vivid example of the problem now confronting China’s AI industry: a model can arrive with an unusually visible development process and still struggle to own the conversation. Its MiMo-V2.6 release included a public training stream, two native multimodal variants, an API route, a desktop client, and MiMo-Code tooling.
Kuaishou’s AI-for-software-engineering group, Kwaipilot, designed KAT-Coder-V2.5 around agents that work in executable codebases, where a suggested fix has to survive files, dependencies, commands, and tests. Kuaishou describes itself as an AI-powered content-community and social-platform company.
Alibaba Cloud’s Qwen team released Qwen3-VL-30B-A3B in October 2025. That older date matters because the model is best read now as a technical backgrounder for a much livelier question: what should a buyer or developer ask before calling a multimodal model a visual agent?
Nikkei Asia has put a hard number on a noisy month for China’s AI industry: its tabulation counted 16 newly developed models launched in September by 10 leading companies, including DeepSeek, Alibaba, Tencent, and Xiaomi. That is a striking pace of activity, but it also exposes the wrong question behind the usual release-race narrative.
Z.ai, the independent Chinese large-language-model provider also known as Zhipu AI, has gained a new route into enterprise software teams: Amazon Bedrock. On Oct. 5, AWS made Z.ai’s GLM-5.3 generally available through the managed AI service. The immediate consequence is practical rather than theatrical.
Kuaishou’s Kling AI is reported to be putting its prospective corporate identity under sharper examination just as its video-generation business begins to produce revenue at a consequential scale.
Beijing E-Town is trying to settle a deceptively basic question about commercial AI personas: when a digital human accepts an order or a virtual gift, who is actually on the other side of the payment? Its answer is to keep the public-facing character and the money-holding entity separate.
South Korea’s Financial Services Commission has made a bank intrusion reported by Shinhan Bank the starting point for a broader test of how the country’s financial system responds when artificial intelligence may be part of an attack.
DeepSeek’s reported new financing round is large enough to shift the conversation from an AI startup’s valuation to the evidence behind its ability to operate at scale.
