DeepSeek raises API prices fourfold with DeepSeek V4 Pro launch

DeepSeek is set to raise its prices significantly with the launch of its latest AI model, DeepSeek V4 Pro, effective August 16. The company plans to increase its API pricing fourfold, with peak hour rates climbing to $3.96 per million output tokens. This marks a dramatic shift for DeepSeek, which has built its reputation on providing lower-cost AI services compared to its Western rivals.

The price hike comes as the company introduces a new pricing structure that differentiates between peak and off-peak hours. During off-peak times, customers will still pay less, at $1.98 per million tokens. The previous pricing of $0.87 for peak hours will no longer be available. DeepSeek's decision to raise prices follows a promotional period that was initially intended to end in May but was extended before the company ultimately chose to increase rates.

Despite the upcoming increases, DeepSeek's new pricing remains competitive within the AI market. For example, the newly priced DeepSeek V4 Flash model will cost $1.32 for peak hours, up from $0.28, while still being more affordable than competitors like Moonshot, whose Kimi K3 model is priced at $15 per million tokens. OpenAI's more advanced GPT-5.6 Sol model is even pricier at $30. However, OpenAI's budget-friendly option, GPT-5.6 Luna, is slightly lower than DeepSeek's peak pricing.

This price adjustment reflects a growing trend in the AI industry where companies are reassessing their pricing strategies amid increasing demand for advanced AI services. As DeepSeek continues to evolve its offerings, it will be crucial for the company to balance affordability with sustainability to maintain its competitive edge in a crowded market.