Skip to main content
GPUBeat Chips & Hardware DeepSeek Slashes V4-Pro AI Model Prices…

DeepSeek Slashes V4-Pro AI Model Prices by 75% Amid Supply Changes

DeepSeek's flagship V4-Pro AI model will see a permanent 75% price reduction, significantly lowering costs for users. This move raises questions about supply chain dynamics in AI chip manufacturing.

In a bold move that reshapes the affordability of advanced AI solutions, DeepSeek has announced a permanent 75% reduction in the price of its flagship V4-Pro AI model. This adjustment positions the V4-Pro at just a quarter of its original price, making high-performance AI technology more accessible to a wider audience.

The price cut lowers API costs to between 0.025 and 6 yuan per million tokens (approximately $0.0035 to $0.83), down from the previous range of 0.1 to 24 yuan. This significant change raises questions about the factors driving this decision. While DeepSeek has not confirmed whether the reduction is linked to an increase in the supply of Huawei's Ascend 950 chips, the implications are noteworthy. The Ascend 950, essential for maximizing the V4-Pro's capabilities, has seen its availability affected by U.S. export controls on competitors like NVIDIA.

The competition for AI technologies is evolving rapidly. Huawei's AI chip sales have gained momentum as U.S. restrictions limit NVIDIA's access to the Chinese market for its most advanced semiconductors. However, similar restrictions on chipmaking equipment exports have constrained Huawei's ability to increase production of the Ascend series. Consequently, while demand for AI solutions is rising, the supply chain dynamics remain complex and challenging.

When DeepSeek introduced the V4 model last month, the Pro version was priced significantly higher—up to twelve times more than its less powerful counterpart, the Flash version. This pricing structure reflected constraints in high-end compute capacity, which limited the availability of the V4-Pro. The company had expected prices to drop sharply once larger quantities of Huawei's supernodes became available later this year, a prediction that now seems to be coming true faster than anticipated.

See also  DeepInfra Secures $107M to Expand AI Inference Cloud Platform

This substantial price cut could have far-reaching consequences for the AI market, potentially prompting other companies to reconsider their pricing strategies. As DeepSeek makes advanced AI technology more affordable, competitors may feel pressured to follow suit, leading to a broader shift in pricing across the industry. This move could accelerate the adoption of AI technologies by small and medium-sized enterprises that previously found such solutions financially prohibitive.

Looking ahead, the future of AI chip production and pricing is likely to remain volatile. The balance between supply constraints and market demand will continue to influence the strategies of AI startups and established players alike. DeepSeek's aggressive pricing may signal the beginning of a new phase in the AI market, where innovation and affordability converge, creating opportunities for a diverse range of users.

As AI technology becomes increasingly integral to various industries, the ripple effects of these pricing strategies will be closely monitored. Companies like DeepSeek are redefining the space, and ongoing developments in chip supply and production will play a key role in shaping the future of the AI crypto market.

Quick answers

What is the new price range for DeepSeek’s V4-Pro model?

The V4-Pro API costs have been reduced to between 0.025 and 6 yuan per million tokens.

Why did DeepSeek implement such a significant price cut?

The company has not disclosed specific reasons, but the availability of Huawei's Ascend 950 chips may be a factor.

GD

GPUBeat Desk

Desk · joined 2026

GPUBeat Desk covers AI infrastructure — chips, foundation models, inference economics, datacenter buildouts, and the geopolitics of compute.