Chinese lab DeepSeek announced the release of a new family of AI models on December 2, 2025: V3.2 and the specialized V3.2-Speciale. According to Bloomberg, the company positions these new releases as direct competitors to top models from Google and OpenAI, claiming parity in reasoning tasks at a significantly lower cost. This release intensifies competitive pressure in the LLM market, demonstrating the rapid progress of Chinese developers and the potential lowering of the price barrier for access to advanced AI.
DeepSeek Launches V3.2 and V3.2-Speciale Models, Challenging GPT-5 and Gemini
