Alibaba Launches Qwen 2.5-Max AI Model, Claims Performance Advantage Over OpenAI and DeepSeek
New AI model integrated into Alibaba’s chatbot aims to challenge leading industry models from OpenAI, DeepSeek, and Meta.
Alibaba has unveiled its latest generative artificial intelligence (GenAI) model, Qwen 2.5-Max, which the company claims outperforms models developed by OpenAI, DeepSeek, and Meta.
According to a statement from Alibaba’s cloud division, the model demonstrates superior capabilities across multiple benchmarks compared to GPT-4o, DeepSeek-V3, and Llama-3.1-405B.
The Qwen 2.5-Max model is already integrated into Alibaba’s AI chatbot, Qwen Chat, and is part of the company's broader efforts to expand its AI research and development.
However, Alibaba has not disclosed the cost of training the new model.
This announcement comes amid intensified competition in the AI industry, particularly following the launch of DeepSeek’s R1 model.
DeepSeek, a Chinese AI startup, recently introduced R1, which has been reported to achieve comparable performance to OpenAI’s models despite being developed with significantly lower computational resources.
DeepSeek claimed that its model was trained with an investment of $6 million in computing power, a fraction of the costs typically associated with large-scale AI training.
The emergence of cost-efficient AI models like DeepSeek’s R1 has led to shifts in the industry, with reports indicating fluctuations in the stock prices of major AI chip suppliers.
Nvidia, a key supplier of AI-related hardware, reportedly experienced a market valuation drop exceeding $500 billion in response to the growing competitiveness of lower-cost AI models, though its stock has since shown signs of recovery.
Meanwhile, concerns have been raised about DeepSeek’s training methodology.
U.S. officials and AI researchers have pointed to evidence suggesting that DeepSeek may have utilized a process known as 'distillation,' a technique where a smaller model is trained using outputs from a larger, more powerful model.
David Sacks, an AI policy advisor in the U.S. administration, stated that there is 'significant evidence' indicating that DeepSeek’s R1 model may have learned from OpenAI’s technology.
OpenAI has not publicly responded to these claims.
As competition in AI development intensifies, leading technology firms continue to invest heavily in refining their models, with Alibaba’s Qwen 2.5-Max representing the latest entry in the global race for AI dominance.
According to a statement from Alibaba’s cloud division, the model demonstrates superior capabilities across multiple benchmarks compared to GPT-4o, DeepSeek-V3, and Llama-3.1-405B.
The Qwen 2.5-Max model is already integrated into Alibaba’s AI chatbot, Qwen Chat, and is part of the company's broader efforts to expand its AI research and development.
However, Alibaba has not disclosed the cost of training the new model.
This announcement comes amid intensified competition in the AI industry, particularly following the launch of DeepSeek’s R1 model.
DeepSeek, a Chinese AI startup, recently introduced R1, which has been reported to achieve comparable performance to OpenAI’s models despite being developed with significantly lower computational resources.
DeepSeek claimed that its model was trained with an investment of $6 million in computing power, a fraction of the costs typically associated with large-scale AI training.
The emergence of cost-efficient AI models like DeepSeek’s R1 has led to shifts in the industry, with reports indicating fluctuations in the stock prices of major AI chip suppliers.
Nvidia, a key supplier of AI-related hardware, reportedly experienced a market valuation drop exceeding $500 billion in response to the growing competitiveness of lower-cost AI models, though its stock has since shown signs of recovery.
Meanwhile, concerns have been raised about DeepSeek’s training methodology.
U.S. officials and AI researchers have pointed to evidence suggesting that DeepSeek may have utilized a process known as 'distillation,' a technique where a smaller model is trained using outputs from a larger, more powerful model.
David Sacks, an AI policy advisor in the U.S. administration, stated that there is 'significant evidence' indicating that DeepSeek’s R1 model may have learned from OpenAI’s technology.
OpenAI has not publicly responded to these claims.
As competition in AI development intensifies, leading technology firms continue to invest heavily in refining their models, with Alibaba’s Qwen 2.5-Max representing the latest entry in the global race for AI dominance.
