Visual China makes its debut with "Investment Account," with MiniMax hidden inside

Ask AI · Why is VIsual China stepping up AI investment despite business downturns under its VIsual China operations?

21st Century Business Herald reporter Lei Chen

Open VIsual China’s recently released 2025 annual report. On one side, revenue and profit from its core business are declining. On the other, the scale of external investments has doubled.

The company’s full-year operating revenue was 778 million yuan, down 4.03% year over year; its net profit attributable to shareholders was 83.0275 million yuan, down 30.31% year over year; and its non-recurring profit was 64.66 million yuan, down 43.02%. The annual report shows that the revenue decline is related to a year-over-year reduction in advertising-industry-related customers’ order volumes.

Meanwhile, cash outflows from the company’s investing activities reached 854 million yuan for the full year, more than double the 386 million yuan in 2024. Most of this money went to wealth management investments. In particular, the investments in MiniMax and Shengshu Technology stand out. MiniMax listed on the Hong Kong Stock Exchange in January 2026, and its market value is over HK$300 billion.

VIsual China’s business model is fairly clear: the upstream connects with more than 800,000 contracted contributors and nearly 300 professional institutions; the downstream serves media, advertising agencies, brand enterprises, and internet platforms, and the company takes a cut from this.

The effectiveness of this platform model hinges on two key metrics: revenue growth rate and gross margin.

In 2025, revenue from the company’s core visual content and services business was 777 million yuan, down 4.11% year over year. In 2024, this revenue was 810 million yuan. The company’s overall gross margin was 41.56%, down more than five percentage points from 46.78% in 2024.

Cost data reveals the reasons. While revenue fell, the company’s operating costs rose 5.22%. In terms of cash flow, net cash from operating activities decreased 34.83% year over year, mainly because the share of project-based businesses increased, leading to higher cost payments associated with them.

When the scale of traditional copyright licensing business shrank, VIsual China’s response strategy was to take multiple routes at once: first, use AI technology to transform traditional copyright transactions, improving search and review efficiency, and develop new features such as intelligent photo editing; second, extend upstream and sell training data to AI large-model companies.

The annual report shows that during the reporting period, to meet core clients’ personalized needs, VIsual China provided content customization services enabled by AIGC for customers in the automotive, fast-moving consumer goods, retail, and media sectors. Content customization services can provide customers with a range of offerings, including AI images, templates, videos, and custom models.

On the customization front, the company’s customized team also collaborated with contracted AI model trainers. Combining high-quality copyright data, they completed training of targeted, customized scenario task models to provide customers with secure and compliant AI studio creative customization services. During the reporting period, the company provided annual creative production services to customers such as Honor, vivo, Mercedes-Benz, Dongfeng Nissan, and Mengniu, and this business grew by more than 18% year over year.

In data services, the company currently has more than 700 million multimodal content data items. It has signed cooperation agreements with companies including Alibaba, Tencent, Meituan, Microsoft, Kuaishou, Datang, Aishi Technology, Honor, and others, to provide data services for large-model training. During the reporting period, the company and its controlling subsidiary Guangchang Creative for the first time carried out revenue-sharing from licensing and authorizing large-model training data to upstream contributors.

In addition, the company is planning to issue H shares and list on the main board of the Hong Kong Stock Exchange. In February 2026, the board of directors has already approved the relevant proposals.

In 2025, VIsual China’s investment actions were nothing short of aggressive. The annual report shows that the investment amount during the reporting period reached 89.7 million yuan, up 266% from 24.49 million yuan in the previous year. These investments are highly concentrated in AI and multimodal technology.

The two most attention-grabbing deals are strategic capital injections into MiniMax and Shengshu Technology. The former is a leading domestic multimodal large-model company; the latter focuses on AI video generation.

VIsual China is not only a financial investor—it is also deeply bound through business synergies. The company’s copyrighted materials have been integrated into Shengshu Technology’s generation platform, along with a compliant content专区 (zone). At the same time, the company’s AI technology platform has also integrated video generation interfaces from MiniMax and Shengshu Technology.

Looking further, VIsual China provides large-model companies with the high-quality, copyright-compliant training data they lack most, in exchange for capabilities to tap cutting-edge generative AI technologies, which in turn supports its own core business. In its annual report, the company summarizes the path of AI capability building as a combination of in-house development, cooperation, and investment.

Its bet on large-model companies has already produced accounting returns. In 2025, VIsual China recognized a fair value change gain of 15.92 million yuan, mainly coming from the valuation increase of its equity in MiniMax. In January 2026, MiniMax listed on the Hong Kong Stock Exchange. As a shareholder, VIsual China’s equity value will be further realized.

The company is trying to build deeper barriers. The annual report proposes a concept: to build an Agent-native intelligent service layer. In other words, VIsual China wants to shift from an image website designed for human users to a foundational data and service interface designed for AI agents.

The shrinking of the old business is a reality; the huge bets aimed at the future are a choice. VIsual China is using today’s profits and cash flow to buy a ticket to the AI world of tomorrow.