MiniMax H1 revenue jumps 283% as enterprise AI demand accelerates

MiniMax Group reported first-half revenue of $116.6 million, up 283.1 percent from a year earlier, powered by a 703 percent jump in enterprise services.

"Our continued investment in model capability, infrastructure efficiency and productization provides the foundation for sustainable growth over the long term," the Shanghai-based company said in its interim results statement.

Revenue from the Open Platform and other AI-based enterprise services climbed 703.1 percent to $73.9 million, accounting for 63.4 percent of total revenue, up from 30.3 percent a year earlier. Revenue from AI-native products, including the Hailuo AI consumer app, rose 100.9 percent to $42.6 million. Gross profit grew 464.8 percent to $20.8 million, with margin expanding to 17.9 percent from 12.1 percent.

The top line still trails the pace needed to hit the $363.77 million full-year consensus compiled by Bloomberg, with the first half representing about 32 percent of that forecast. Shares closed up 1.13 percent at HK$303 on Wednesday ahead of the announcement.

Net loss narrowed 11 percent to $358 million from $402 million a year earlier, though adjusted net loss widened 111.2 percent to $293 million after adding back share-based payments, fair-value losses on financial liabilities and listing expenses. Research and development spending rose 138.8 percent to $296.9 million, a slower pace than revenue growth, as the company trained its M-series foundation models.

International markets contributed about 60.8 percent of total revenue, reflecting what MiniMax calls a "global-first commercialization strategy" spanning more than 230 countries and regions. The company serves more than 300 million individual users and over one million enterprises and developers.

MiniMax, founded in December 2021 and backed by Alibaba Group Holding Ltd. and Tencent Holdings Ltd., completed its Hong Kong IPO in January 2026. It competes with Z.AI Co. and other Chinese AI labs racing to release top-tier large language models. During the period it released the MiniMax M3 model, strengthening coding and agentic capabilities, and followed with the open-weights MiniMax H3 video-generation model after the reporting period.

The results show MiniMax converting model advances into paid enterprise workloads faster than its consumer apps, a shift that should support margin expansion as inference demand grows. Investors will watch the company's conference call on Wednesday evening for updated guidance on full-year revenue and enterprise adoption.

This article is for informational purposes only and does not constitute investment advice.