DeepSeek to increase API prices following surge in model demand

DeepSeek has announced upcoming significant price increases for its API services. The decision follows the rapid global adoption of its DeepSeek-V4-Flash-0731 model, which has become the fastest-growing model ever measured in terms of token usage on the Ollama platform.

Industry analysts suggest the price hike may serve as a traffic regulation tool to manage extreme service loads caused by the model's high performance-to-cost ratio. While DeepSeek has built its reputation on being significantly cheaper than competitors—with some tasks costing roughly 105 times less than Anthropic’s Claude models—the company faces the challenge of maintaining stability amidst intense competition.

Because DeepSeek has released its model weights under a license that allows for redistribution, third-party providers such as DeepInfra and OpenRouter are able to host the model on their own hardware. This allows some third-party platforms to offer rates lower than DeepSeek’s official API prices. However, developers have noted that official APIs often provide better efficiency, such as higher cache hit rates, which can impact the actual cost-effectiveness of third-party subscriptions.