<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/63ec66ba-416a-4e91-93ec-7c5a8345483d.jpeg?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/63ec66ba-416a-4e91-93ec-7c5a8345483d.jpeg"/>
As one of the most 关注 ed AI model companies in the Hong Kong stock market this year, MiniMax has delivered its first annual report card.
Along with the financial report, it showcased its ambition to transition from a &#34;model company&#34; to an &#34;AI platform company.&#34; Against the backdrop of intensified global competition in large models and the capital market gradually shifting from &#34;storytelling&#34; to &#34;seeing results,&#34; this financial report clearly holds significant directional meaning.
According to the financial report, MiniMax achieved rapid revenue growth in 2025. Revenue from AI-native products surged from $21.805 million in 2024 to $53.075 million, a year-on-year increase of 143.4%. This is mainly attributed to the viral reputation of products like HaiLuo AI and Talkie/Xingye globally, as well as a significant increase in user willingness to pay.
At the same time, revenue from the open platform and enterprise services aimed at B-end customers also skyrocketed by 197.8% to $25.963 million, proving the scale effect of its model API in enterprise-level applications. The change in revenue structure indicates that MiniMax is moving away from reliance on single model call revenue, gradually forming a dual-driven business pattern of C-end products and B-end services.
Against the backdrop of soaring revenue, MiniMax's gross profit achieved a leap of 437.2%, reaching $20.079 million, with the gross profit margin doubling from 12.2% to 25.4%.
Industry insiders believe that MiniMax has not been caught up in the old playbook of &#34;burning money for traffic&#34; from the traditional internet era, successfully establishing a path for natural growth driven by product strength.
MiniMax COO Eric Wu told Wall Street Insights that while major domestic companies may be most concerned about DAU, MiniMax is more focused on whether it can define a new generation of intelligence.
The product head of MiniMax also stated, &#34;AI is an era where production determines demand. What should we look at? The simplest is to see what models users are willing to pay for.&#34;
Technological advancements are another main thread of this financial report.
In Q4 2025, MiniMax consecutively launched the M2, M2.1, and M2-her models, and further released the M2.5 version in February 2026. Data shows that the average daily token consumption of the M2 series text models in February 2026 has increased more than six times compared to December 2025, with consumption related to Coding Plan increasing more than tenfold.
The explosive growth in token consumption indicates that the enhancement of model capabilities is driving real demand expansion.
Company founder Yan Junjie proposed that in the coming year, the intelligence level of models will further leap, with the programming field entering the L4 to L5 level of intelligence, moving from &#34;tools&#34; to &#34;colleague-level&#34; collaboration; office scenarios will replicate last year's progress speed in the programming field; and multi-modal creation will achieve &#34;directly deliverable&#34; medium to long content, even leading to new forms of real-time streaming output
In his view, the combination of these three factors will lead to a large-scale explosion of intelligent supply and open an unprecedented window of innovation at the application layer. This judgment is both a prediction of technological trends and a declaration of the company's strategic direction.
For MiniMax, last year was a decisive year for crossing from single-modal to full-modal integration.
The video model HaiLuo 2.3 released last year significantly improved in action performance and image quality, with the cost of batch creation reduced by up to 50%. By the end of the year, it had generated over 600 million videos; the speech model Speech 2.6 supports more than 40 languages and achieves ultra-low latency interaction, generating over 200 million hours of speech; the music model Music 2.5 realized more complex emotional expressions and longer output durations.
The synergy of the four modalities—language, video, speech, and music—constitutes its &#34;full-modal&#34; technology matrix and provides underlying support for products like HaiLuo AI.
In addition to external product iterations, MiniMax also conducted an experiment at the organizational level. The company disclosed that its internal Agent interns have covered nearly 90% of employee positions, introducing AI collaboration from programming development to market sales. In early 2026, the company will productize its internal capabilities and launch the MiniMax Agent AI-native Workspace.
This &#34;first for internal use, then for external sale&#34; approach positions the company as a testing ground for AI-native organizations, attempting to feed product innovation through improved organizational efficiency. If this model proves successful, MiniMax will not only be a model supplier but may also become an infrastructure provider for enterprise-level AI productivity platforms.
Overall, this annual report conveys three key signals. First, the commercialization path is gradually becoming clear, and revenue growth is sustainable. Second, gross margins have significantly improved, and technological efficiency is beginning to translate into financial efficiency. Third, model capabilities are in a rapid leap phase and are extending towards multi-modal and Agent directions.
Of course, challenges still exist.
High R&amp;D investments will continue to suppress profit performance, and fluctuations in global computing costs and international competitive pressures cannot be ignored, with the profitability timeline still unclear. In 2025, MiniMax's R&amp;D expenditure reached $2.53 billion, an increase of 33.8% year-on-year. Increased investment in cloud services and computing training led to an adjusted net loss of $250 million in 2025.
In the capital market, as a star company in the Hong Kong stock AI sector, MiniMax's valuation has partially reflected high growth expectations. Whether it can continue to deliver on technological leadership and business expansion will be the core variable determining stock price performance.
From a long-term perspective, MiniMax is at a critical juncture. If programming capabilities from L4 to L5 are truly realized in 2026, and office and multi-modal content achieve large-scale commercial use, the company's transition from a &#34;large model developer&#34; to a &#34;platform company&#34; will have a realistic foundation. At that time, what it carries will no longer be just model invocation demands, but may become the foundation of a new generation of productivity systems.
The performance in 2025 is a capability verification; the execution in 2026 will determine whether MiniMax can truly complete this strategic upgrade

00100

- MiniMax, a prominent AI modeling company, reported significant revenue growth in 2025, with AI product revenue soaring by 143.4% to $53.1 million.  
- The company's shift from a model-focused approach to an AI platform model is evident as it seeks to support both consumer and enterprise services, leading to a notable increase in gross profit margins.  
- Although challenges such as high R&D investments and market competition persist, MiniMax's strategic evolution towards becoming a comprehensive AI productivity platform could redefine its operational foundation in the coming years.

When the model begins to grow into a platform: MiniMax's turning point