Cheap AI Models from China Attract Global Users and Reshape the Stock Market

Chinese wave of cheap AI models is turning into a full-fledged economic phenomenon: new players are capturing global audience, and investors are frantically searching for the next winners on the Chinese stock market. Bloomberg reports: the token-economy revolution is already reshaping the balance of power in the global artificial intelligence industry. After it became clear to the world in early 2025 that a competitive large language model could be trained many times cheaper than American counterparts, Chinese tech companies received a powerful signal: price war is their main weapon.

Dozens of models with aggressive pricing appeared on the market: Alibaba through Qwen, Baidu with ERNIE, ByteDance with its own developments, as well as startups like Zhipu AI and Moonshot. The cost of generating a million tokens from Chinese providers is often 5–20 times lower than from OpenAI or Anthropic. It is from this price dynamics that the concept of token-economy grew.

Its essence is simple: when inference becomes cheaper, the number of requests grows exponentially. Developers who previously couldn't afford AI integration into their product due to high API costs are now launching applications, agents, automations. The market expands not at the expense of richer users, but at the expense of everyone else — and Chinese providers were the first to ride this wave.

Global reach is not just a marketing thesis. According to analysts, Chinese models are actively used in Southeast Asia, India, Latin America and Arab countries, where price sensitivity is particularly high. APIs from Alibaba Cloud or Baidu cost literally pennies compared to Western alternatives, and the quality gap has practically disappeared.

Companies are adapting models for local languages, which further accelerates penetration. Stock market reaction turned out to be quite expected. Shares of companies that either develop their own models or build infrastructure for their delivery are sharply revalued.

Particularly notable gains among cloud computing suppliers, among corporations that have embedded AI functions into mass-market products — from e-commerce to educational platforms. The Chinese analogue of the "Magnificent 7" appears to be already forming, though its composition has not yet settled. In parallel, the geopolitical dimension is intensifying.

The US continues to tighten export controls on high-performance chips, hoping to slow Chinese development. However, the effect turns out to be ambiguous: restrictions force Chinese companies to develop more efficient algorithms and architectures. A deficit of computational resources has turned into a stimulus for engineering ingenuity.

The American companies' monopoly on frontier model development is gradually being destroyed — not because someone created something revolutionary, but because price itself has become a competitive advantage. For businesses around the world, there is now a real choice between expensive Western APIs and cheap but fully productive Chinese counterparts. The question is no longer who has the smartest model — the question is who will first build the infrastructure to deliver intelligence at prices that will make it accessible to the next billion users.