DeepSeek plans mid-July V4 release with 2x peak-hour API pricing

Midday and afternoon peak pricing may push global users to shift workloads, influencing deployment choices and cost controls around the V4 launch.

Summary

DeepSeek told developers in an upgrade email that its official DeepSeek V4 release is planned for mid-July and will introduce peak and off-peak API pricing. Standard rates for V4 Pro and V4 Flash will remain unchanged outside peak periods, but prices will double from 09:00-12:00 and 14:00-18:00 daily. During those windows, V4 Pro output will cost 12 yuan per million tokens and V4 Flash output will cost 4 yuan per million tokens. The pricing structure points to a demand-management approach often used to balance heavy usage at busier times, and it may prompt global users to adjust usage patterns as they manage AI deployment costs.

Terms & Concepts
  • API: A way for developers to access a model or software service programmatically.
  • tokens: Units of text that AI models process for input and output billing.
  • peak and off-peak API pricing: A billing model that charges higher rates during busier time windows and standard rates at other times.