The Chinese AI Startup Deepsek has not yet determined the time of its R2 model’s release as CEO Liang Wenfeng is not satisfied with his performance, the report reported on Thursday told two people citing the knowledge of the situation.
R2One successor Deepsek Wildly popular R1 The Reasoning model, was planned for release in May, with goals to produce better coding and cause in languages beyond English, the Reuters reported earlier this year.
For the past several months, the engineers of Deepsek have not been working to refine R2 unless the Leiang gives the green light for the release.
However, the US export regulations can result in rapid adoption of R2 due to lack of NVIDIA server chips in China, the report states that the ventures that enter the enterprises citing the employees of the top Chinese cloud firms, which provide the models of the lampsac to the venture customers.
A potential increase in R2 demand will overwhel Chinese cloud providers who need advanced Nvidia Chips to run the AI model, the report states.
Deepsek did not immediately respond to the request of a Reuters for comments.
The report stated that Deepsek is in touch with some Chinese cloud companies, which offers them technical specifications to direct their plans for hosting and distribution of models from their servers.
Currently in its cloud customers using R1, the majority are running models with NVidia’s H20 chips, stated in the information.
In April, fresh export curbs imposed by the Trump administration prevented Nvidia from selling in their H20 chips in the Chinese market – the only AI processor can be legally exported to the country at that time.
© Thomson Reuters 2025
(This story is not edited by NDTV employees and auto-generated from a syndicated feed.)