- Microsoft has announced that Depseek R1 will arrive at Co -Machines+ Machines
- It will be executed on the device, so it will be less powerful than other versions of the model.
- Deepseek R1 will have three versions of co -pilot+ that will be launched over time
Deepseek has seriously shaken the world of AI with a LLM that is apparently cheaper to train, more efficient in power and, nevertheless, equally intelligent compared to its rivals. While Google, Open AI and others fight to decipher how Depseek’s model became so impressive from nothing, with Openai even stating that he copied chatgpt to get there, Microsoft is taking the ‘if you can’t beat them They, join them ‘, join them’ ‘approach in their place.
Microsoft has announced that, after the arrival of Deepseek R1 in Azure Ai Foundry, you can soon execute an Optimized version of Deepseek’s AI in its co -pilot+ PC. This feature will be extended first to Qualcomm Snapdragon X machines, followed by Intel Core Ultra 200V laptops and AI AMD chips.
It will begin by making Depseek-R1-Distill-qwen-1.5b available in Microsoft AI Takit for developers, before unlocking the most powerful versions of 7b and 14b. While these are not as impressive as the 32B and 70b variants also at their disposal, versions 14b and seniors of Deepseek can work on the device.
This mitigates one of the main concerns with Depseek: that data shared with AI could end in unusual foreign servers, with Microsoft and added that “Deepseek R1 has suffered rigorous red and safety equipment evaluations” to further reduce possible risks security
How to get Depseek R1 in co -pilot+
To start using Copilot+ Build DePseek on a device once it is available, you will need an Azure account: you can register on the official Microsoft website if you still do not have one. His next step will be to start Azure Ai Foundry and seek Deepseek R1. Then press
After a few moments, the chat games option must open, and you can start chatting with Depseek on the device.
If you have not yet used Deepseek, two great advantages that you will find when you install it is that it is currently free (at least for now), and that you show your “thought” as you develop your answers. Another AI, like Chatgpt, goes through the same thought process, but they don’t show it, which means that you have to refine your indications through a test and error process until you get what you want. Because you can see your process, and where it could have come out on the wrong way, you can modify your Speeek indications to achieve your goals.
Like the 7b and 14B variants unlock, it should see the Azure model of Deepseek R1 improve, although if you want to try it, you may want to do it sooner rather than later. Given the serious Microsoft association with Openai, we hope that this emerging rival does not try if it turns United, Italy and other regions.
