Zhipu and MiniMax are quickly becoming two of the most watched names in China’s fast-moving AI services market. Their revenue is climbing at an eye-catching pace, helped by surging demand for generative AI tools across businesses and everyday users. But there’s a catch: their losses are widening just as fast. That combination is a major warning sign for where the industry is headed, because it suggests 2026 could bring a harsh shakeout for Chinese AI service startups that don’t have enough cash, scale, or a clear route to profitability.
What’s happening is straightforward but brutal. AI model startups are finally seeing meaningful commercialization momentum, with customers paying for AI-driven products and services in increasing numbers. At the same time, building and operating competitive AI models remains extremely expensive. These companies must keep training, fine-tuning, and optimizing their models to stay relevant, while also running stable commercial operations that meet customer expectations. This “two-front war” pushes research and development spending up, increases operating costs, and accelerates cash burn.
The result is that even rapid top-line growth may not be enough to keep a company safe. Operating losses can expand so quickly that available capital starts shrinking at an alarming rate. When that happens, the risk isn’t only lower valuations or slower expansion—it’s the very real possibility of cash-flow disruption, where a business struggles to fund ongoing compute needs, staffing, and daily operations. In an AI services market where performance improvements and product reliability are make-or-break, any interruption can quickly lead to customer churn and stalled momentum.
Zhipu and MiniMax illustrate how different go-to-market strategies can still run into the same financial pressure.
Zhipu’s approach emphasizes combining and deploying AI model capabilities in ways that fit enterprise needs. That positions it strongly for business AI services, where customers may pay more for dependable deployments, customization, and integration into workflows. The upside is potentially higher-value contracts and longer relationships. The challenge is that enterprise sales cycles can be demanding, service expectations are high, and supporting business clients adds operational complexity—often requiring significant upfront investment before profitability shows up.
MiniMax leans more toward AI quality and consumer-facing applications, aiming for broad user adoption and global reach. Consumer AI services can scale quickly when a product hits the right fit, but the economics can be unforgiving. Large user bases are expensive to serve, and the gap between “many users” and “many paying users” can be wide. Even when monetization improves, infrastructure costs can climb alongside usage, keeping margins under pressure.
This is why investors are becoming more pointed in what they want to see. The market is shifting from excitement over AI potential to scrutiny of revenue quality and, increasingly, profitability. Growth still matters, but it’s no longer enough on its own. Startups are being pushed to prove that revenue can outpace the rising costs of compute, talent, model iteration, and customer support.
Looking toward 2026, the broader expectation is that AI services will commercialize faster, with more industries adopting AI tools and more consumers relying on AI-enabled apps. Paradoxically, that acceleration could make the market tougher, not easier. As competition intensifies, companies will feel even more pressure to spend on model capability, performance, and reliability—while customers demand better results at lower prices.
That’s why the concept of a “startup death valley” is becoming a common way to describe the next phase. It’s the point where product demand is real, but the cost of staying competitive is so high that only the best-capitalized, best-managed, and best-positioned players can keep going. For China’s AI service startup scene, 2026 may be the year when growth stories face their hardest test: turning momentum into sustainable business.
