Woofun AI reports that DeepSeek has notified developers of an upcoming pricing adjustment for its V4 API, scheduled for mid-July release. The new structure introduces a peak-valley model where prices double during designated high-demand periods.

Peak hours are defined as 9:00 AM to 12:00 PM and 2:00 PM to 6:00 PM. During these windows, V4 Pro cache hits cost 0.05 yuan per million tokens, cache misses 6 yuan, and output 12 yuan. V4 Flash peak rates are 0.04 yuan for cache hits, 2 yuan for cache misses, and 4 yuan for output. Regular prices remain unchanged outside these hours.