Alibaba releases AI model it claims surpasses DeepSeek-V3

Key Points

  • Alibaba released Qwen 2.5-Max, claiming it outperforms DeepSeek-V3 and other leading AI models.
  • The release timing on Lunar New Year highlights the competitive pressure from DeepSeek's recent successes.
  • DeepSeek's low-cost AI models have caused a price war and prompted upgrades among Chinese tech companies.
  • DeepSeek's founder, Liang Wenfeng, focuses on achieving AGI rather than engaging in price wars.

Summary

Alibaba has launched Qwen 2.5-Max, an AI model that it claims outperforms several leading AI models including DeepSeek-V3, GPT-4o, and Llama-3.1-405B. The release on the first day of the Lunar New Year underscores the competitive pressure exerted by DeepSeek, a Chinese AI startup whose recent advancements have not only shaken Silicon Valley but also spurred a rush among domestic competitors to enhance their AI offerings. DeepSeek's low-cost models have initiated a price war in China, with companies like Alibaba, Baidu, and Tencent slashing prices to remain competitive. Despite this, DeepSeek's founder, Liang Wenfeng, emphasizes a focus on achieving artificial general intelligence (AGI) over engaging in price wars, highlighting a contrast between DeepSeek's lean, research-focused operation and the more traditional, large-scale structures of China's tech giants.

yahoo
January 29, 2025
Stocks
Read article

Related news