Microsoft and Hugging Face expand collaboration; Azure AI Studio adds 1-click deploy for Llama 3, Mistral 7B, Qwen 1.5, and Phi-3; AMD MI300X VMs announced

AI Impact Summary

Microsoft and Hugging Face announce a broad collaboration to bring open models to enterprise on Azure, enabling one-click deployment from the Hugging Face Collection into Azure AI Studio and Model Catalog. The rollout includes popular open models (Llama 3, Mistral 7B, Qwen 1.5 110B, Command R Plus, Phi-3) and new hardware support with AMD MI300X-based ND VMs, plus on-device/local inference via ONNX Runtime Web/WebGPU. This shifts how teams prototype and deploy open-model workloads, offering faster start times, improved cost/performance on Azure, and deeper integration with Spaces Dev Mode and VS Code for developers.