As interest mounts regarding a new version of the DeepSeek AI model, various Chinese tech firms have launched their own generative artificial intelligence models recently. Notable among these are Alibaba’s Qwen 3.5 and ByteDance’s Seedance 2.0. However, MiniMax, which debuted in Hong Kong in January, has attracted particular attention from AI users and UBS stock analysts alike. In mid-February, MiniMax released its M2.5 model, showcasing performance comparable to Claude’s Opus 4.6 but at a significantly lower cost, drawing a growing number of developers toward it.
UBS analysts reported that MiniMax’s AI usage had reached one-third of Anthropic’s Claude at just one-tenth of its price, prompting the firm to initiate coverage with a buy rating and a price target of 1,000 Hong Kong dollars (approximately $127.83), representing a projected increase of over 30% from its last trading price of HKD763.50. Analysts expressed confidence in MiniMax’s potential to capture market share in the global enterprise sector, largely due to its diverse suite of AI tools that extend beyond text generation to include video and audio facilitation.
In contrast, competitors like Zhipu and Moonshot have different focuses, such as coding-related functionalities. The anticipation surrounding the new DeepSeek model has sparked heightened interest among Chinese internet companies, which have sought to enhance customer engagement during the recent Lunar New Year holiday. By incorporating AI capabilities into apps, these companies aim to expand AI application among users, particularly in less developed regions.
UBS estimates suggest MiniMax could potentially capture 3% of the global enterprise services market, translating to a revenue opportunity of around $41 billion. If its AI leadership advances more rapidly than currently projected, UBS forecasts the stock price could rise significantly.
Why this story matters
Key takeaway
Opposing viewpoint
