/pcq/media/media_files/2025/01/31/H6wqbyCunA3yC8riErqL.png)
The AI battle in China is getting intense. During the Lunar New Year, Alibaba released an AI model—Qwen 2.5-Max. The release timing is likely due to the pressure from DeepSeek, an AI startup that is rising fast and is going to shake up the domestic and global markets with its recent progress.
Alibaba claims Qwen 2.5-Max beats OpenAI’s GPT-4o, Meta Llama 3.1 405B, and DeepSeek-V3 model in most AI benchmarks. And this is another big step for Alibaba to catch up with the western AI giants in China.
DeepSeek and Its Worldview
Just this month, DeepSeek was exposed and launched DeepSeek-V3, followed by R1 on January 20. With reports showing the two systems can do things as complex as OpenAI’s bare cost, the situation got intense in the competitive battlefield. They moved so fast that the stocks in Silicon Valley started to wobble as investors wondered if the top U.S. AI players could afford the cost of these large models.
The open-source pricing started the price war in AI among Chinese companies since DeepSeek launched DeepSeek-V2 with a price of 1 yuan ($0.14) for 1 million tokens and Alibaba lowered its price by 97%. Baidu and Tencent will follow.
Alibaba's Qwen 2.5-Max: A Response to DeepSeek?
With the rapid rise of DeepSeek, Alibaba's AI strategy appears to be shifting. Qwen 2.5-Max is positioned as a direct counterattack, with Alibaba claiming its performance outpaces DeepSeek-V3 across multiple key benchmarks.
Key Differences: Alibaba AI vs. DeepSeek
Feature |
Alibaba AI |
DeepSeek AI |
---|---|---|
Focus Area |
Enterprise AI, Cloud Services |
AGI Research, Open Source AI |
Business Integration |
Deeply embedded in Alibaba Cloud |
Independent, research-driven |
Pricing Strategy |
Competitive, after DeepSeek’s price war |
Extremely low-cost models |
Regulatory Ties |
Strong ties with Chinese regulators |
Potential government scrutiny |
Innovation Speed |
Balances business and research |
Prioritizes rapid AI breakthroughs |
AI War Very Fierce in China
Meanwhile, Alibaba's recent maneuvers symbolize the rivalry among China's tech giants for the top spot in AI in both domestic and global arenas. Competitors like ByteDance are also ramping up. Only days after DeepSeek announced its R1 model, ByteDance announced its updated AI, which outperformed OpenAI's benchmarks.
With all the monetary backing for AI, DeepSeek approaches things very differently. It is more like a research lab under the mysterious Lien Wen Feng, who only wishes to attract young and fresh Ph.D. talents from universities of global admiration and is not at all inclined to have anything resembling top-down orders flowing from some boardroom.
Liang expressed his opposition to pricing wars in an interview, acknowledging that the ultimate goal of DeepSeek was "AGI," artificial general intelligence, which is an AI that could outperform human beings on most tasks deemed of economic value.
Global AI Markets Expectations
The Alibaba vs. DeepSeek spat shows the global AI landscape. Chinese companies and American companies are no longer competing with each other; they are racing with each other at lightning speed.
With this AI rush, here are the burning questions:
-
Is startup-focused and lean DeepSeek trying to take on corporate giants like Alibaba?
-
What will the US tech giants do about the AI coming out of China?
-
Will the AI price war in China make AI affordable for the world or expose other sustainability barriers?
/pcq/media/media_files/2025/01/31/oRgnxOIBdgf1nKTidmZR.png)
What does it mean for China’s AI future?
Alibaba’s Qwen 2.5-Max marks a new chapter in the China-AI war, where even big tech companies are aware of the intensifying competition in this space against other fierce startups like DeepSeek. These rivalries bring about faster changes—in technology and business as well as the global power structure.
The duel? One thing is for sure: the AI war is far away from being done.
Also Read:
Top 5 GPT Tools Revolutionizing Cybersecurity
ChatGPT vs. DeepSeek vs. Gemini : Which AI Model Fits Your Needs?