Starting Wednesday, users will be able to access the DeepSeek V3 model at half its regular price from 12:30 a.m. to 8:30 a.m. Beijing time. Additionally, the R1 model, which powers the company's chatbot, will see a price reduction of 75%.
This initiative comes as the Hangzhou-based artificial intelligence startup faces overwhelming interest and demand from users and developers embracing its technology. Competitors, including Tencent Holdings Ltd. and Perplexity AI, have begun utilizing DeepSeek’s models, alongside various state entities in China, such as departments within the Hong Kong government.
On Tuesday, DeepSeek resumed credit top-ups for API access, although it has warned that capacity will be limited during daytime hours.
Since its launch on January 20, DeepSeek’s reasoning chatbot has challenged traditional beliefs regarding the costs associated with developing advanced AI models. It has been recognized as a competitor to OpenAI while requiring significantly lower development expenses. This has led to a surge in Chinese tech and internet stocks, as investors anticipate further growth for these companies, despite facing restrictions from U.S. chip trade policies.