DeepSeek-V4-Pro announces that API prices are permanently reduced to a quarter of the original price. Will this spark another price war among models?

robot
Abstract generation in progress

[Caixin] On May 22, DeepSeek announced on its website that the API price for the DeepSeek-V4-Pro model will be officially adjusted to one-quarter of the original price. This price discount was originally offered as a limited-time promotion, set to expire on May 31.

After the price adjustment, the DeepSeek-V4-Pro will cost 0.025 yuan per million tokens of input (cache hit), 3 yuan for input (cache miss), and 6 yuan for output.

This makes the API price of the DeepSeek-V4-Pro model lower than that of mainstream large models worldwide, making the API price competitive. Taking input price as an example, the input (cache miss) price for DeepSeek-V4-Pro is 3 yuan per million tokens, approximately $0.44; GPT-5.5, Claude Opus 4.7, and Claude Opus 4.6 are all priced at $5, GPT-5.4 at $2.5, Gemini 3.1 Pro at the lowest $2, Grok 4.3 at $1.25, and another domestic large model Kimi K2.6 at $0.95.

View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments