DeepSeek V4 launching in July with surge pricing for API calls

DeepSeek just announced that the official release of DeepSeek V4 is scheduled for mid-July, and it's coming with a twist: surge pricing. Under the new peak/off-peak pricing model, API calls made during Beijing time daily from 9:00 AM to 12:00 PM and 2:00 PM to 6:00 PM will cost twice the usual rate.

Here's how the pricing shakes out for the high-performance deepseek-v4-pro model. During off-peak hours, you'll pay 0.025 yuan per million tokens for input when the cache hits, 3.00 yuan when it misses, and 6.00 yuan for output. During peak hours, those prices double to 0.05 yuan, 6.00 yuan, and 12.00 yuan respectively. For the lighter deepseek-v4-flash model, off-peak pricing is 0.02 yuan for cache-hit input, 1.00 yuan for cache-miss input, and 2.00 yuan for output. Peak prices jump to 0.04 yuan, 2.00 yuan, and 4.00 yuan.

DeepSeek says it will notify users via email 24 hours before any pricing change takes effect.