According to PerceptOne AI's monitoring, DeepSeek officially announced that DeepSeek V4's stable version is scheduled to be launched in mid-July, with the simultaneous introduction of the Peak-Valley Pricing Mechanism. During the peak hours of 9:00 to 12:00 and 14:00 to 18:00 Beijing time, the API billing price will be adjusted to twice the weekday rate.
Under the new pricing mechanism, the high-performance model deepseek-v4-pro will have a weekday price of 0.025 yuan for each million tokens for input cache hits, 3.00 yuan for cache misses, and 6.00 yuan for output. During peak hours, these prices will be adjusted to 0.05 yuan, 6.00 yuan, and 12.00 yuan, respectively. The lightweight model deepseek-v4-flash will have a weekday price of 0.02 yuan for input cache hits, 1.00 yuan for cache misses, and 2.00 yuan for output per million tokens. During peak hours, these prices will be adjusted to 0.04 yuan, 2.00 yuan, and 4.00 yuan. Users will be informed via email 24 hours before the actual billing adjustment.
