Microsoft’s New Deal With Mistral AI
By Adam Pease
Microsoft’s New Deal With Mistral AI
Microsoft has just announced a partnership with French AI provider Mistral.
This blog discusses Mistral and the implications of the deal for the generative AI market.
Who Is Mistral?
Mistral is a French AI model provider largely known for its release of open source large language models (LLMs). Mistral’s LLMs have scored competitively on a variety of benchmarks, offering even the leading productized models, like ChatGPT, a run for their money.
While Mistral began as an open-source model provider, some of its recent moves appear to suggest a pivot similar to the product-oriented direction of OpenAI. Mistral’s new, larger models are no longer open-sourced. It will be interesting to see whether Mistral continues to lean in this direction as it scales in partnership with Microsoft, or if it attempts to balance open-source research with proprietary products.
The Microsoft Partnership
The new partnership will bring the underlying infrastructure power of the Microsoft Azure compute cloud to Mistral’s models. As researchers struggle to accelerate the performance of compute-hungry LLMs, startups that want to build foundation models from the ground up, like Mistral, must find ways to support the large GPU requirements.
The partnership itself makes Mistral’s model services available through the Azure AI services cloud, announcing research and scaling of AI systems as parallel goals. With this move, Microsoft cements its leadership as an AI model provider. The move may even invite regulatory scrutiny, as Microsoft now arguably controls the distribution of both the state-of-the-art closed-source provider (OpenAI) and the state-of-the-art closed-source provider (Mistral).
Bottom Line
The deal benefits Microsoft’s consolidation of AI services under Azure, while also providing Mistral the opportunity to bootstrap its systems to greater scale in a time of scarce compute.
Get Ready to Transform With Us for Our First Transform Tour Stop of 2024!
Going LIVE on Thursday, March 15th, 2024 at 10 AM PT | 1 PM ET
Since the launch of OpenAI’s GPT service, organizations have been scrambling to put out their own copilot service.
In our March Transform Tour, taking place on Thursday, March 14, 2024, we will cover “The Battle of Edge vs. SaaS Computing” and “Putting CoPilots to Work in 2024 and 2025.”
Have a Comment on this?