Huawei Cloud and top Chinese LLMs launch 'Hundred Models' AI ecosystem

Huawei Cloud’s INSPIRE announcement is a clear bid to cement a homegrown alternative to the US‑centric model oligopoly. Rather than pushing a single giant model, Huawei is positioning itself as the cloud and orchestration layer for an ecosystem of Chinese LLMs—Zhipu’s GLM‑5, DeepSeek, MiniMax, Kimi, StepFun, Baidu’s Ernie, iFlytek Spark, Meituan’s vertical models, and more. The message is that enterprises can mix and match specialized models on Huawei Cloud, with commercial, regulatory, and data‑sovereignty guarantees tailored to the mainland market. Strategically, this is about reducing dependence on US chips and cloud while still moving fast on agentic and multimodal capabilities. If Huawei succeeds in making it easy for Chinese developers to deploy, route, and monetise dozens of domestic models through a common platform, it will intensify the regional arms race in tooling and applications rather than just raw model size. Over time, that could feed back into core research: the more production traffic flows through these systems, the more training signal Chinese labs have to push their next‑generation models closer to frontier levels.