Microsoft integrates Grok 3 AI into Azure for health and science use

USA – Microsoft has announced the integration of Grok 3, the flagship AI model from Elon Musk’s xAI, into its Azure platform.

The move is part of Microsoft’s AI Foundry initiative, which supports developers in building, managing, and deploying AI applications at an enterprise level.

Grok 3 and a smaller version, Grok 3 Mini, will now be available on Azure. These models were developed using xAI’s Colossus supercomputer, reportedly offering 10 times more computing power than earlier models.

The goal is to push the boundaries of artificial intelligence, particularly in fields such as healthcare and scientific research.

According to Microsoft,“This collaboration combines xAI’s cutting-edge models with Azure’s enterprise-ready infrastructure, giving developers access to Grok 3’s advanced capabilities in a secure, scalable environment.”

The model excels in fields such as mathematics, coding, reasoning, and world knowledge, with a deep expertise in sectors like science and medicine.

Grok 3 has been designed to aid in medical diagnoses and support scientific research workflows. Microsoft and xAI have both emphasized its value in healthcare.

The model is available through a free preview for two weeks, after which pricing starts at US$ 3 per million input tokens and US$ 15 per million output tokens for Grok 3 Global. Grok 3 (DataZone) will cost slightly more. It can also be accessed via GitHub.

Microsoft’s Foundry platform also features AI models from major developers like OpenAI, Meta, NVIDIA, and Cohere.

This partnership with xAI aims to expand the range of intelligent tools available to organizations operating in high-impact fields.

In November, Elon Musk encouraged users on X (formerly Twitter) to “try submitting X-ray, PET, MRI or other medical images to Grok for analysis.”

He acknowledged that the feature is still in its early stages but claimed it is “already quite accurate and will become extremely good.” He also asked users to provide feedback on Grok’s performance to help improve the model.

However, Grok’s approach to medical image interpretation has raised concerns. Privacy regulators in Europe are investigating how xAI handles personal data.

The Irish Data Protection Commission (DPC) launched an inquiry into whether the AI model is trained using public posts on X, which could violate GDPR rules.

xAI initially introduced the Grok-1 chatbot in November 2023, calling it “anti-woke” and giving it a humorous, bold personality.

“Grok is designed to answer questions with a bit of wit and has a rebellious streak, so please don’t use it if you hate humor,” the company noted in its launch statement.