MiniMax’s latest financial results show a Chinese AI company becoming more dependent on enterprise customers even as it continues to spend heavily on the models and infrastructure behind its products. The company reported first-half revenue of $116.6 million, up 283.1% from a year earlier, while its Open Platform and other enterprise AI services expanded much faster than its consumer-facing products.
Reuters reported that revenue from the enterprise segment reached $73.9 million, a 703.1% year-on-year increase. That business represented 63.4% of MiniMax’s revenue in the first six months of 2026, compared with 30.3% in the same period a year earlier. The shift gives a clearer view of where the company is finding paying demand: less in novelty consumer tools and more in services that other businesses can build into their own workflows.
The result does not mean the startup has left the investment-heavy stage of AI behind. Its loss attributable for the period narrowed to $358 million from $402.2 million a year earlier. But the financial mix matters because China’s foundation-model sector increasingly has to demonstrate that demand can become recurring revenue rather than free usage, investor enthusiasm, or benchmark visibility.
MiniMax has already appeared in EastFrontier coverage through its model releases, including the open-weight H3 video model. The new filing moves the story from model capabilities to commercial traction. It suggests that the company’s economic case may rest more on an enterprise platform than on any one consumer application.
Enterprise Services Become the Core of MiniMax’s Revenue Mix
MiniMax’s Open Platform and enterprise AI services produced $73.9 million in the first half. That figure is not merely a large growth rate from a small base. It made the segment the company’s largest source of revenue. The 63.4% contribution share shows that enterprise demand now outweighs the consumer AI-native products that helped establish MiniMax as one of China’s prominent model startups.
The company’s consumer AI-native products still grew quickly, rising 100.9% to $42.6 million. Yet the contrast between the two lines is revealing. Consumer products may produce visibility, data, and a direct relationship with users. Enterprise services can produce more predictable contracts, API consumption, and software integrations. In a market where many models offer similar basic chat and generation features, that difference can determine who turns usage into a durable business.
MiniMax said it will continue pursuing what it calls the “performance-cost frontier,” meaning it wants models to handle demanding real-world tasks while becoming inexpensive enough to deploy more broadly. That is the company’s description of its strategy, but it captures the commercial pressure facing Chinese AI firms. Customers do not necessarily need the most expensive available model for every task. They need systems that produce useful output at a price compatible with production use.
This is one reason enterprise adoption has become central to the sector. A company may accept some technical trade-offs if a model can be operated at a lower cost, connect to existing software, and scale across many employees or customers. For MiniMax, the steep rise in enterprise-service revenue shows this proposition is gaining commercial acceptance, even if it remains unclear how durable those customers will be as rivals cut prices.
Fast Revenue Growth Has Not Eliminated the Cost of Scale
The company’s losses narrowed, but they remained substantial at $358 million for the half. That tension is normal for a model company that must pay for training, inference, cloud capacity, research talent, and product development before revenue fully catches up. It is also why headline growth should not be read as a simple measure of financial maturity.
MiniMax has raised capital to support that expansion. Reuters noted that it had previously completed a Hong Kong listing and raised additional funding through a share sale and bond issue. Those financing steps give it more room to invest, but they also raise expectations that its commercial strategy can show a route to improving returns.
The company is not alone in facing that challenge. China’s AI market has a large population of providers promising lower-priced models, open weights, agents, and specialized tools. Earlier EastFrontier reporting on China’s AI video lead showed how MiniMax competes in an ecosystem where model builders are constantly seeking adjacent applications. Financial results provide a different measure of the race: whether the underlying systems are being bought often enough to support ongoing investment.
The first-half data indicates that MiniMax’s enterprise business is helping answer that question. A 703.1% increase in the segment cannot by itself establish a long-term trend. Revenue can be influenced by the timing of large contracts, platform launches, or customer migrations. Still, the shift from 30.3% to 63.4% of company revenue is material enough to show that the firm is no longer relying primarily on consumer products.
China’s AI Pricing Contest Moves Into the Enterprise Stack
The broader implication is that China’s AI competition is increasingly taking place inside enterprise stacks. The public contest over benchmark scores and model size remains important, but enterprise customers buy access, reliability, integration, and cost control. They may use a platform to power customer service, generate content, build internal tools, or support developers, none of which requires the same model configuration.
That setting rewards providers that can package models into a practical service layer. MiniMax’s Open Platform is one such layer. Its growth indicates that businesses are willing to pay for access to AI capabilities without taking on the full burden of hosting and operating a model themselves.
The company’s H1 results therefore tell two stories at once. The encouraging story is that enterprise AI services now account for most of revenue and are growing rapidly. The harder story is that the company is still losing hundreds of millions of dollars while it pursues scale. Both are necessary to understand the state of China’s AI startup economy.
MiniMax has not yet demonstrated that a large enterprise segment automatically leads to profitability. It has demonstrated something more immediate: businesses are becoming the center of its revenue model. As Chinese model providers compete on price and deployment, that may prove more consequential than the next viral consumer feature.