Chinese artificial intelligence startup DeepSeek is raising API prices for its V4-Pro and V4-Flash models, introducing a new pricing structure that varies depending on usage time and token type.

The updated rates will take effect on August 17, with increases ranging from approximately 50% to as much as 1,100% compared with current prices, depending on the model and type of API request.

DeepSeek will also introduce peak and off-peak pricing, allowing API costs to change depending on when developers access its models. The move could help the company manage computing demand while improving revenue from its rapidly expanding AI services.

The pricing change represents a notable shift for DeepSeek, which has gained attention partly by offering capable AI models at significantly lower costs than many Western competitors.

As demand for advanced AI models increases, infrastructure and computing costs are becoming increasingly important factors in how AI companies price access to their technology.