LLMPrice
Updated daily
DeepSeek · Direct API

DeepSeek V4 Flash pricing

Peak rates are shown by default. DeepSeek charges exactly half outside its published weekday UTC peak windows.

DirectOfficial provider
Input$0.44per 1M tokens
Cached input$0.014per 1M tokens
Output$1.32per 1M tokens
Context1Mmaximum window
Endpoint and source

Direct API

Open Official provider source ↗

This is a first-party rate checked against the model provider's official pricing page.

Caching
2,000 input + 500 output tokens$0.00154 / request
Calculate a complete workload →