Chinese artificial intelligence startup DeepSeek announced a sweeping reduction in its model pricing, cutting the price of input cache hits across its entire API suite to one-tenth of prior levels effective immediately. The company said the change is designed to speed developer adoption and reduce computing expenses for users of its services.
In addition to the permanent price adjustment for input cache hits, DeepSeek said its recently launched V4-Pro model will continue to be offered at a 75% discount through May 5. The firm described the extension of that promotional pricing as part of an aggressive push following the model's rollout earlier this month.
DeepSeek's moves come amid intensifying competition among global providers of AI services, with vendors competing to attract developers and enterprise customers by delivering cheaper, more efficient large language models. Companies in the sector have increasingly treated lower inference and usage costs as a central battleground as they work to expand adoption beyond early experiments.
Observers note that DeepSeek's pricing step is likely to increase pressure on rival providers to re-examine their own pricing frameworks, particularly among firms operating in China. That market has seen a rapid buildout of domestic alternatives, a trend that has accelerated in the context of U.S. technology restrictions.
DeepSeek framed its price changes as a way to lower barriers to use for developers and to reduce the compute costs that can make scaling applications expensive. The extended promotional pricing on V4-Pro follows the model's initial launch earlier this month and keeps the discounted rate in place through the specified May 5 date.
While the company emphasized immediate cost reductions for users of its API suite, the broader market implications include intensified focus on pricing from other providers and continued competition on model efficiency and operational cost. The move underscores how vendor pricing strategies are now a central element of efforts to drive broader enterprise and developer adoption of large language models.
Clear summary
DeepSeek has cut input cache hit prices across its API suite to one-tenth of previous levels effective immediately and extended a 75% discount on its V4-Pro model until May 5, part of an aggressive push to attract developers and lower compute costs amid rising competition in the AI services market.