Do you want to fly high or far? DeepSeek and ByteDance are the leaders in China’s AI industryDifferent strategies are adopted by.
DeepSeek V3.2 released Monday is another version of DeepSeek. open-weight model Anyone can play around with it. The startup claims that it is on par or even better than the latest OpenAI models and Google on key math benchmarks.
ByteDance – a software company whose dominance in AI applications We covered earlier, introduce even more Doubao, its chatbot is available in a variety of ways. ByteDance has teamed up with a Chinese manufacturer of smartphones to integrate Doubao in the operating system. This will give it access to various apps, and allow it to perform agentic tasks. Also, Apple’s Siri will be affected.
ByteDance’s and DeepSeek’s AI apps are used by over 140,000,000 people each month. However, their recent announcements show two divergent trends in China’s AI industry. Some Chinese companies continue to compete against their Western counterparts in order to develop ever more advanced models. However, other firms have quietly abandoned this competition and focussed on how AI can help people.
DeepSeek Resurfaces
DeepSeek’s newest open-weight models may disappoint some of their most loyal customers, as they are still anticipating R2, the much-anticipated upgrade to the original model which rocked Silicon Valley last January. DeepSeek instead released V3.2, and V3.2 Speciale, better optimized versions of the previous model V3.2 Exp released in September.
V3.2 has caused quite a stir among AI experts because DeepSeek is said to be able to solve questions from the International Mathematical Olympiad. Its performance in other tasks such as coding or reasoning should also match, if not surpass, GPT 5 or Gemini 3. “It suddenly dawned on me why they call the company DeepSeek with the whale as a motif. Because just like a whale, it rarely surfaces, but every time it surfaces, it always makes a massive splash,” Jen Zhu Scott is an AI investor, cofounder, and CEO at Power Dynamics. The firm offers modular data center solutions.
This AI arms race is becoming a bit tiresome, especially because there have been so many models released over the past month that each claim to be able to move humanity forward. OpenAI’s GPT 5.1 was released in just 20 days. Google’s Gemini 3 Pro followed, as did Anthropic’s Claude Opus 4.5. Add to that Chinese open source models, such Moonshot’s Kimi K2 and DeepSeek’s V3.2. The sum of my attention span is this perfect meme.
“At the end of the day, we can’t keep up with all these hairline differences between different models, different releases,” Zhu says. “It actually doesn’t make a huge difference, apart from some kind of stock market speculation on who’s gonna win.”

