DeepSeek announces flash series AI model price cut for tomorrow

On September 9, 2026, DeepSeek announced that it will implement new pricing for its Flash series AI models starting at 12:00 Beijing Time on September 10, 2026. The updated pricing includes an output rate of 4 yuan per million tokens. During off-peak hours, input cache hits are priced at 0.02 yuan per million tokens, while input cache misses are priced at 1 yuan per million tokens. Peak hour rates are double those of off-peak hours. The Flash series includes two models, deepseek-v4-flash and deepseek-v4-flash-vision-exp, with price reductions ranging from 11.11% to 60% across various token types and usage periods.