Chinese tech behemoth Alibaba has introduced its Qwen 2.5-Max AI model, boldly asserting that it has outpaced DeepSeek’s renowned DeepSeek-V3 model, reports First Post.
Unveiled on the first day of the Lunar New Year, the launch of Qwen 2.5-Max highlights the intensifying competition within China’s AI sector, reflecting the pressure DeepSeek’s rapid ascent has exerted not only on global competitors but also on domestic ones, UNB reports.
Alibaba's cloud division revealed on WeChat that Qwen 2.5-Max outperformed OpenAI’s GPT-4, DeepSeek-V3, and Meta’s Llama-3.1-405B across various performance metrics. The timing of this announcement, during the Lunar New Year festivities, underscores the urgency felt by Chinese firms to maintain competitiveness against DeepSeek, which has made waves in the AI market since its January debut.
DeepSeek’s Market Disruption
DeepSeek’s sudden success, beginning with the launch of its AI assistant powered by the DeepSeek-V3 model on January 10 and followed by the R1 model on January 20, has disrupted the tech industry. The Chinese startup’s cost-effective approach to developing powerful AI has raised concerns in Silicon Valley, particularly as investors question the high development costs associated with leading US companies. In response, Chinese competitors are racing to enhance their models.
ByteDance, for example, updated its flagship AI model shortly after DeepSeek’s R1 release, claiming it surpassed OpenAI’s GPT-1 in the AIME benchmark test, which assesses AI’s ability to comprehend and respond to complex commands. This move highlights how DeepSeek’s swift rise has spurred action among domestic firms, with Alibaba’s latest release being a response to DeepSeek’s innovations.
The Emergence of a New Competitor: Kimi k1.5 from Moonshot
Complicating the race further is Moonshot AI’s new Kimi k1.5 model, which launched just days after DeepSeek’s R1. Kimi k1.5 is being regarded as a direct rival to both DeepSeek’s models and OpenAI’s GPT-4, with reports suggesting it outperforms both on key benchmarks. Unlike DeepSeek-R1, which lacks multimodal capabilities, Kimi k1.5 is a multimodal model capable of processing and reasoning across text, images, and code, giving it a substantial advantage for tasks requiring both visual and textual data.
Kimi k1.5 has also been developed at a fraction of the cost compared to similar cutting-edge AI models in the US, positioning Moonshot AI as a growing force in the global AI arena. Its advanced reinforcement learning techniques further enhance its versatility, making it highly adaptable to a range of applications.
How DeepSeek stacks up against ChatGPT and Gemini
Shifting Dynamics in China’s AI Industry
China’s rising influence in AI is becoming increasingly apparent, as companies like DeepSeek, Alibaba, and Moonshot AI challenge the longstanding dominance of US tech giants. The launch of DeepSeek-V2 last May ignited an AI price war in China, prompting Alibaba’s cloud division to reduce prices by up to 97%. This price-cutting strategy has since become common practice among Chinese firms, including Baidu and Tencent, as they strive to develop AI models that can compete with OpenAI and other US-based giants.
DeepSeek, led by Liang Wenfeng, has taken a distinct approach, operating more like a research lab with a lean team of graduates and PhD students. Liang’s vision of achieving artificial general intelligence (AGI) with significantly lower overhead than larger tech companies contrasts with the more costly, hierarchical models of Alibaba and other Chinese tech giants.
As China’s AI sector continues to evolve at a rapid pace, its influence on the global market grows. The competition between DeepSeek, Moonshot AI, and Alibaba marks a crucial shift in the AI landscape, with these startups and tech giants pushing the boundaries of AI development. The race for AI supremacy is underway, and China is leading the way.
