The commercial trajectory of China’s artificial intelligence sector is accelerating faster than Wall Street previously anticipated, driven by a combination of aggressive pricing strategies and rapid technical breakthroughs. According to a report by the South China Morning Post, Goldman Sachs has revised its run-rate revenue forecast for the Chinese AI model market upward by 30 percent, pushing the projected total to US$13 billion.
The updated forecast, published on Monday, marks a significant jump from the bank’s previous US$10 billion estimate and underscores the intense commercial momentum building within the domestic ecosystem.
This upward revision is fundamentally rooted in the shifting economics of intelligence. As Chinese developers release increasingly capable foundation models at a fraction of the cost of their Western counterparts, corporate adoption is accelerating. Goldman Sachs noted that the competition for the best performance-to-price balance is intensifying among Chinese AI models, creating a highly favorable environment for enterprise integration.
This dynamic is not only expanding the total addressable market but also driving significant revenue growth for the leading model developers who are successfully capturing global token share.
The Impact of Open-Weight Pricing on China’s AI Market
The catalyst for this accelerated revenue growth is the fierce price war currently defining the Chinese AI landscape. The Goldman Sachs report specifically highlighted the recent launch of MiniMax’s H3 model, an open-weight system capable of processing text, image, video, audio, and music. By pricing the H3 model at just 30 to 50 percent of incumbent market levels, MiniMax has significantly lowered the barrier to entry for developers and enterprises seeking multimodal capabilities. This aggressive pricing strategy is forcing competitors to respond, driving down the overall cost of intelligence and stimulating broader market demand.
In response to this momentum, Goldman Sachs lifted its projected year-end annualized recurring revenue (ARR) for MiniMax to US$1 billion. The bank also raised its year-end ARR projection for Zhipu AI, another leading domestic developer, to US$2.5 billion. These substantial revenue targets indicate that Chinese AI startups are successfully transitioning from the research and development phase to large-scale commercial monetization. By offering highly capable models at disruptive price points, these companies are not only capturing domestic market share but also positioning themselves as formidable competitors on the global stage.
DeepSeek V4 Flash and the Coding Benchmark Race
Beyond pricing, the technical capabilities of Chinese models are also advancing rapidly, further justifying the upgraded revenue forecasts. The Goldman Sachs note highlighted the official launch of application programming interface (API) access for DeepSeek’s V4 Flash model, which occurred on the same day as the MiniMax H3 release. DeepSeek’s latest offering has achieved front-end coding capabilities that rival Zhipu AI’s flagship GLM-5.2 model, demonstrating the rapid pace of iteration within the domestic ecosystem.
According to the latest data from the Arena AI leaderboard, Zhipu’s GLM-5.2 currently ranks seventh globally in front-end coding performance, with DeepSeek’s V4 Flash closely following in eighth place. This level of technical proficiency, combined with highly competitive API pricing, makes these models highly attractive to software developers and enterprise IT departments.
As more companies integrate these AI tools into their software development workflows, API and subscription revenues are expected to surge. Summaries of the Goldman Sachs note indicate that total revenues from China’s large AI models could grow from an estimated 35 billion yuan in 2026 to a staggering 879 billion yuan in the coming years, highlighting the immense long-term commercial potential of the sector.
(Related: Goldman Sachs Urges Investors to Go Long on China’s AI Value Chain Amid Triple Mismatch)
Strategic Implications for the Global AI Market
The upgraded revenue forecast from Goldman Sachs carries significant implications for the broader global AI market. As Chinese developers prove their ability to generate billions of dollars in annualized recurring revenue, they will have the financial resources necessary to sustain massive investments in compute infrastructure and next-generation model training. This financial independence is critical for maintaining competitiveness against well-funded Western rivals like OpenAI and Anthropic.
Furthermore, the success of the Chinese open-weight and low-cost API model challenges the prevailing commercial strategies in Silicon Valley, where frontier models are often kept closed and priced at a premium. If Chinese companies can continue to deliver near-frontier performance at a fraction of the cost, they may successfully capture a significant portion of the global enterprise market, particularly in cost-sensitive regions across the Global South. The US$13 billion revenue projection is not just a measure of domestic success; it is a clear indicator that China’s AI sector is building a sustainable, highly lucrative commercial engine capable of powering its long-term technological ambitions.
The Goldman Sachs revision also carries important implications for how investors should think about the competitive dynamics between Chinese and American AI developers. The bank’s analysis suggests that the most consequential battleground is no longer raw model capability but rather the ability to deliver consistent, reliable performance at the lowest possible cost per token. Chinese developers have proven adept at this optimization challenge, leveraging architectural innovations like Mixture-of-Experts (MoE) and advanced distillation techniques to extract maximum performance from constrained hardware budgets.
This efficiency-first approach is not merely a workaround for US chip export controls; it has become a genuine competitive advantage that is reshaping the economics of the entire industry. As Goldman Sachs lifts its revenue projections, it is implicitly acknowledging that the era of American dominance in commercial AI monetization is over, replaced by a genuinely bipolar market where Chinese and US developers compete on roughly equal footing for the loyalty of global enterprise customers.