(Reuters) – Microsoft has made the DEEPSEEK Chinese Startup artificial R1 model available on its Azure Cloud Computing platform and its GitHub tool for developers, the American company announced on Wednesday.
The AI model will be available in the catalog of models on platforms and will join more than 1,800 models that Microsoft offers.
Deepseek launched a free AI assistant last week who, according to him, uses less data to a fraction of the cost of the services in place. On Monday, the assistant exceeded American rival downloads in the Apple App Store downloads, causing panic among investors in technological action.
This decision comes as Microsoft sought to reduce his dependence on Chatgpt Maker Openai. The company worked to add internal and third -party AI models to supply its flagship AI Microsoft 365 COPILOT product, Reuters reported last month.
Microsoft also said that customers could soon run the R1 model locally on their co -pilot + PCS, a decision that could potentially facilitate confidentiality and data sharing concerns about the use of the model.
Deepseek said that he stores user information in servers in China, which could be a point of collision in its American adoption.
Meanwhile, Microsoft and Openai have investigated whether OpenAi technology data data was obtained unauthorized by a group linked to Deepseek, Bloomberg News reported on Tuesday.
Deepseek exploding on the AI scene has encouraged the rivals to answer, the boss of Openai, Sam Altman, saying that the company “will draw some outings” – after which he published Tuesday a personalized version of Chatgpt for government agencies American.
Alibaba in China also published a new version of its Qwen 2.5 AI model on Wednesday, an unusual timing, considering that it marked the first day of the new lunar year.
(Report by Deborah Sophia in Bengaluru; edition by Maju Samuel)