DeepSeek quadruples V4 model API prices
TL;DR. DeepSeek announced significant price increases for its V4 Pro and V4 Flash API models, effective August 16. - DeepSeek V4 Pro peak hour costs will rise from $0.87 to $3.96 per million output tokens. - The company is implementing a new peak and off-peak pricing structure to manage resource allocation. - Despite the price hikes, DeepSeek's V4 models remain more affordable than many Western competitors.
- DeepSeek is increasing API prices for its V4 Pro and V4 Flash models by up to four times.
- New pricing includes peak and off-peak rates, with peak hours costing $3.96 per million output tokens for V4 Pro.
- The company previously offered promotional pricing which was briefly made permanent before the current hike.
- DeepSeek's new prices, while higher, still position its models as more affordable than some rivals like OpenAI's GPT-5.6 Sol or Moonshot's Kimi K3.
Sources
- DeepSeek's AI models are about to cost four times more — engadget.com