BRIEFLY.
DeepSeek Permanently Cuts V4-Pro AI Model Costs by 75%
1 min read
Briefly Editorial Team

DeepSeek Permanently Cuts V4-Pro AI Model Costs by 75%

TL;DR

  • 75% discount on V4-Pro is now permanent
  • Huawei chip optimization reduced costs
  • Company competes with Western AI giants

Why it matters

DeepSeek's decision strengthens Chinese AI companies amid Western sanctions against Nvidia and opens new opportunities for corporate users.

Key Details of the Discount

Chinese AI startup DeepSeek announced a permanent 75% discount on its flagship V4-Pro model, launched in 2026. The API cost with the discount ranges from $0.0035 to $0.83 per 1 million tokens. This decision stems from optimizing the model to run on Huawei Ascend 950 chips, significantly lowering operational costs.

Strategy Amid Sanctions

U.S. export restrictions on Nvidia GPUs have forced Chinese companies to seek alternatives. DeepSeek's shift to Huawei chips reduced costs and enabled more competitive pricing. Experts believe this move will strengthen Chinese AI companies in the global market.

Competition with Western Giants

The V4-Pro discount aims to attract developers and corporate clients. By leveraging local technology, DeepSeek competes with Western AI giants, a critical advantage as access to Western accelerators is restricted. This strategy is vital for maintaining market relevance in a sanctions-driven environment.

Market Prospects

U.S. export controls on Nvidia have created a market boom for Huawei and other Chinese chipmakers. DeepSeek capitalizes on this by expanding its market share and reinforcing its position in AI services. Lower prices could stimulate innovation in Chinese firms and accelerate domestic AI industry growth.