Models & Pricing | DeepSeek API Docs
The prices listed below are in units of per 1M tokens. A token, the smallest unit of text that the model recognizes, can be a word, a number, or even a punctuation mark. We will bill based on the total number of input and output tokens by the model.
AI Summary
DeepSeek offers two main API models: deepseek-v4-flash and deepseek-v4-pro, both supporting thinking and non-thinking modes with a context length up to 1M tokens and maximum output of 384K tokens. Pricing is calculated per 1M tokens, with deepseek-v4-flash costing $0.435 for input (cache miss), $0.003625 for input (cache hit), and $0.87 for output tokens, while deepseek-v4-pro is currently offered at a 75% discount until May 31, 2026. Input cache hit prices were recently reduced to 1/10 of their launch price effective April 26, 2026, and billing is based on total input and output tokens consumed.