Microsoft has announced that you will soon be able to use DeepSeek on Copilot+ PCs.
The DeepSeek R1 model is casting a high light. Its low development cost surprised many large tech companies, and OpenAI suspects that ScaleAI used data from GPT-4 to train its model. After the R1 model, similar AI models such as Dall-E and Stable Diffusion also launched. Microsoft, on the other hand, apparently believes in the power of the AI model, and will soon offer it on Copilot+ PCs.
Quick move
In a blog, Microsoft announced that DeepSeek will be rolled out first to Qualcomm Snapdragon X PCs, followed by Intel Lunar Lake and AMD Ryzen AI 9 PCs. The DeepSeek-R1-Distill-Qwen-1.5B model will be available “soon” in the Microsoft AI Toolkit for developers. Later, the more powerful 7B and 14B variants will appear.
The 1.5B base model is not powerful compared to the 32B and 70B models, but all three are “NPU-optimized” for Copilot+ PCs. The R1 model is already available on the Azure AI Foundry platform, reports The Verge.
DeepSeek may deliver strong performance, but the developers at ScaleAI are clearly less knowledgeable about security. A database containing DeepSeek’s sensitive information was openly exposed on the Internet. That database is now said to be back under lock and key, but the news is not very positive for the Chinese start-up.