Microsoft has announced that it will bring a distilled version of DeepSeek R1 AI models to Windows 11 Copilot+ PCs. The announcement comes shortly after the availability of DeepSeek R1 on Azure AI Foundry, Microsoft’s developer focussed platform that aims to drive innovation using AI tools and ML models in a responsible and secure manner. The integration will first arrive on Snapdragon X powered laptops, tablets and PCs, with Lunar Lake and AMD Ryzen AI 300 processors following suit at a later date.
Through a post on its Windows Developer Blog, Microsoft shared that Neural Processing Unit (NPU) optimized versions of DeepSeek-R1 are coming to Copilot+ PCs. This would include thin and light laptops from all major manufacturers including Asus, Lenovo, Dell, HP, and Acer; all of them sell systems powered by Snapdragon X Plus and X Elite processors with 45 TOPS NPUs. Copilot+ PCs all have NPUs with 40+ TOPS.
The first release which is titled DeepSeek-R1-Distill-Qwen-1.5B will be available to developers via the AI Toolkit and soon after, the 7B and 14B versions will become available as well. Microsoft says these optimized models will “let developers build and deploy AI-powered applications that run efficiently on-device, taking full advantage of the powerful NPUs in Copilot+ PCs”.
While the company has not shared an exact timeline for the release, Intel Core Ultra 200V devices are set to get these new powers faster than their AMD Ryzen AI counterparts. In the meantime, AMD has already shared DeepSeek R1 Distill-related guides for its latest Ryzen AI 300 series APUs, as well as for Ryzen 8040/7040 chips despite the latter featuring just 16/10 TOPS NPUs, respectively.