Alibaba Unveils Its Most Powerful AI Model While DeepSeek's V4-Flash Becomes the Most Affordable

Chinese company Alibaba unveiled its largest and most powerful AI model to date on Monday, leading to a sharp rise in its stock price. In the meantime, DeepSeek's latest AI model is reported to be more than 100 times cheaper than Anthropic's Claude Fable 5. These two events underscore the rapid pace of AI development by Chinese tech companies, which are fiercely competing to create more powerful systems without making them prohibitively expensive to run. Both models, Alibaba's Qwen3.8-Max and DeepSeek's V4-Flash, demonstrate China's commitment to open-source models, with companies aiming to gain popularity among developers worldwide. An analyst from research firm Omdia noted that Chinese AI companies have found an important market, as many business processes require models that are good enough, affordable, transparent, and easy to use, and open-source models help meet this demand. Alibaba's Qwen3.8-Max soared in rankings after its debut, leading to a 7% jump in the company's stock price in Hong Kong. The new model supports 2.4 trillion parameters, slightly behind its domestic competitor Kimi K3 from Moonshot AI, which launched last month with 2.8 trillion parameters. Qwen3.8-Max topped the Arena.AI leaderboard for Chinese models in the text category but still trails behind Claude Fable 5 and three Opus variants from Anthropic. For vision models, it ranks second globally, just behind Claude Fable 5. Both Qwen3.8-Max and Kimi K3 can process text, images, and video with a context of up to 1 million tokens. Alibaba stated that Qwen3.8-Max, scheduled for release next week, uses a 'mixture of experts' architecture, simultaneously utilizing 95 billion parameters to reduce costs and latency. According to research company Artificial Analysis, DeepSeek's V4-Flash, released on Friday, is the least expensive model among known models worldwide, costing $0.14 per million input tokens and $0.28 per million output tokens, with an average test cost of $0.03 compared to $0.86 for Kimi K3, $1.86 for GPT-5.6 Sol from OpenAI, and $3.15 for Claude Fable 5.