Microsoft Expands GitHub and Azure AI Foundry with the DeepSeek-R1 AI Model
On Wednesday, Microsoft made its Azure AI Foundry and GitHub compatible with the recently published DeepSeek-R1 artificial intelligence (AI) model. These platforms now provide access to the reasoning-focused AI model. The software behemoth from Redmond also emphasized that it has carried out thorough safety assessments to guarantee that the model is secure for personal and business use. Through the AI Toolkit, the business is also providing Copilot+ PC customers with condensed versions of the R1 model.
Microsoft Introduces DeepSeek-R1 to GitHub and Azure AI Foundry
The IT giant said in a blog post that the DeepSeek-R1 AI model is now accessible through GitHub and Azure AI Foundry’s model catalogue. Notably, Azure AI Foundry is an enterprise-focused platform that allows developers to create, test, and develop custom copilots and generative AI applications.
“One of the main benefits of using DeepSeek R1 or any other model on Azure AI Foundry is the speed at which developers can experiment, iterate, and integrate AI into their workflows,” stated Asha Sharma, Corporate Vice President of AI Platform at Microsoft, who made the announcement.
The article also mentioned that the business performed safety assessments and red teaming on the R1 model, which included thorough security checks and automated model behavior evaluations. According to Microsoft, it has taken steps to reduce any such dangers. Additionally, Azure AI Foundry offers extra security through its Safety Evaluation mechanism and Content Safety filtering mechanism.
Users can look for DeepSeek-R1 in the model catalogue to get started. Once they have located it, they must click on deploy and open the model card. The inference application programming interface (API) needed to execute the model will be made available to the user.
Distilled DeepSeek-R1 Will Be Available to Users of Copilot+ PCs
Microsoft said that it is introducing optimized versions of DeepSeek-R1 to Copilot+ PCs in a different post. These purified versions will be added to the Intel Core Ultra 200V chipset and other devices after initially appearing on Snapdragon X-powered devices. The DeepSeek-R1-Distill-Qwen-1.5B model, which is available through the AI Toolkit, will be the first model to be released. Later on, the 7B and 14B versions will be included.
