China’s AI startup DeepSeek is offering developers a huge 75% discount on its newly released AI model, DeepSeek-V4-Pro, until May 5. The company also announced deep price cuts for its entire DeepSeek API lineup.
In a post on X, DeepSeek stated that it is also lowering prices for input cache hits across its entire DeepSeek API lineup to just one-tenth of the original cost. These aggressive pricing strategies aim to attract more developers and increase the adoption of its AI models.
On Friday, DeepSeek launched a preview of its highly anticipated V4 model, which has been specially adapted for Huawei’s chip technology. This adaptation highlights China’s focus on developing autonomous AI solutions using domestic hardware.
The V4 model comes in two versions: the more powerful and higher-priced “Pro” version, and the lighter, more affordable “Flash” variant. This tiered offering allows developers to choose the model that best fits their specific needs and budget.
DeepSeek claims that the Pro version outperforms other open-source models in benchmarks related to world knowledge, only falling behind Google’s closed-source Gemini-Pro-3.1. This positions DeepSeek-V4-Pro as a very competitive option in the AI market.
According to the Chinese startup, the V4 models are especially well-suited for “AI agent work.” AI agents can perform more complex tasks than regular chatbots, but they require much greater computing power. DeepSeek’s focus on this area suggests a push towards more sophisticated, autonomous AI applications. These significant price reductions will likely make its advanced AI technology more accessible to a wider range of developers, fostering innovation and competition in the AI landscape.
