DeepSeek provides powerful AI models accessible through their API. Understanding their pricing structure is crucial for effective usage and budget management. This article delves into the specifics of DeepSeek API's pricing model, including token costs, context caching, and deduction rules, providing you with a clear understanding of how you'll be charged.
DeepSeek API pricing is based on the number of tokens processed. A token is the smallest unit the AI model recognizes, which can be a word, number, or punctuation mark. You are billed for both input tokens (the data you send to the API) and output tokens (the data the API returns). It's essential to understand token usage to optimize your API calls and control costs. For more detailed information about token and token usage, refer to Token & Token Usage.
Here's a breakdown of the pricing for DeepSeek's primary models:
Key Considerations:
deepseek-chat model has been upgraded to DeepSeek-V3, and deepseek-reasoner now points to the new model DeepSeek-R1.deepseek-reasoner model uses CoT to provide reasoning steps before the final answer. All tokens from CoT and the final answer are included in the output token count and priced equally. You can learn more about CoT reasoning using the Reasoning Model guide.max_tokens isn't specified in your API request, the default maximum output length is 4K tokens. Adjust this parameter to support longer outputs as needed.DeepSeek offers context caching to potentially reduce costs. With context caching, if the model can reuse previously processed information, the input token price is significantly lower (the "Cache Hit" price). Check DeepSeek Context Caching for more details and how to leverage this feature.
The total cost of using the DeepSeek API is calculated using a simple formula:
Expense = (Number of Tokens) x (Price per Token)
Fees are deducted directly from your account balance. If you have both a topped-up balance and a granted balance, the granted balance will be used first.
DeepSeek reserves the right to adjust product prices. It's recommended to:
For further information and updates, you can connect with DeepSeek through the following channels:
By understanding these pricing details, you can effectively utilize the DeepSeek API while managing your budget and optimizing your AI-powered applications.
