A year after DeepSeek shook global markets with a low-cost artificial intelligence model, Chinese rivals are preparing a new wave of releases aimed at matching or surpassing its impact. The Hangzhou-based firm’s breakthrough in early 2025 accelerated China’s shift toward affordable, open-source AI systems, challenging assumptions that only massive infrastructure spending could deliver top-tier performance.
Now, several companies are gearing up to unveil upgraded models around China’s Spring Festival period. Zhipu AI has launched a new model emphasizing coding and long-duration task execution. ByteDance introduced Seedance 2.0, a video-generation model positioned for high-end content creation, while also preparing upgrades to its Doubao chatbot. Meanwhile, DeepSeek is expected to release its next-generation V4 model, and Alibaba is reportedly close to unveiling its Qwen 3.5 series.
DeepSeek’s earlier release triggered a global tech selloff and erased nearly $600 billion from Nvidia’s market value in a single session. Its models significantly undercut Western pricing, with research indicating Chinese systems operate at a fraction of comparable U.S. costs. The combination of open-source access, lower deployment expenses, and competitive reasoning capabilities has since become a defining strategy among Chinese AI developers.
Unlike some rivals facing investor pressure to commercialize quickly, DeepSeek remains structurally insulated, backed by a quantitative hedge fund. Competitors such as Alibaba are increasingly focusing on integrating AI into consumer-facing applications, including e-commerce features embedded in chat interfaces.
As anticipation builds, analysts say expectations are high. The coming launches may determine whether China’s AI sector can sustain its momentum—and whether low-cost, open-source models remain the dominant approach in the next phase of global AI competition.
