DeepSeek raises API prices fourfold with V4 Pro model

DeepSeek will raise its API prices fourfold on August 16, 2026, with the launch of its V4 Pro and V4 Flash models.

The new pricing introduces peak and off-peak rates to "allocate resources more reasonably."

V4 Pro will cost $3.96 per 1 million output tokens during peak hours, up from $0.87, and $1.98 off-peak. V4 Flash will cost $1.32 per 1 million output tokens during peak hours, up from $0.28, and $0.66 off-peak.

The reversal undoes a May 2026 announcement that the promotional rates, originally set to end on May 31, 2026, would become permanent.

Even after the increase, DeepSeek's V4 models remain cheaper than many competitors. Kimi K3 from Moonshot costs $15 per 1 million output tokens, OpenAI's GPT-5.6 Sol costs $30, and OpenAI's GPT-5.6 Luna costs $1.20, below the V4 Flash peak-hour rate of $1.32.

DeepSeek built its reputation on offering AI services at a fraction of the cost of Western rivals; the new pricing marks a shift as it seeks to manage resource allocation more effectively with its latest models.