What is Azure Mistral and How Does it Work

Azure Mistral is a cloud-based platform that provides a secure and scalable way to manage and analyze data. It's a part of the Azure family of cloud services.

Azure Mistral is built on top of the Apache Airflow platform, which is an open-source workflow management system. This means that Azure Mistral inherits many of the features and capabilities of Apache Airflow.

The platform is designed to help users manage complex workflows and data pipelines in a scalable and reliable way. It allows users to create, manage, and monitor workflows that can handle large volumes of data.

Azure Mistral provides a user-friendly interface for creating and managing workflows, making it accessible to users without extensive technical expertise.

Mistral Large (2402), Mistral Large (2407), Mistral Small, and Mistral Nemo can be deployed as a serverless API with pay-as-you-go billing.

To deploy a Mistral model, start by opening Azure AI Studio and navigating to the Model Catalog.

You can find the Mistral-large model by searching for it or by clicking on the "View Models" under the Mistral model.

To deploy the model, click on "Deploy" and select the "Pay-as-you-go" option.

Choose the project in which you want to deploy your model, and make sure it's in the East US 2 or France Central regions.

The pricing of the Mistral Large model based on tokens can be found in the Marketplace Offer Details section.

Each project has its own subscription to the particular Azure Marketplace offering of the model, which allows you to control and monitor spending.

You can only have one deployment for each model within a project.

Give the deployment a name, which will become part of the deployment API URL.

The deployment API URL and Secret Key can be found on the Deployments page after the deployment is ready.

You can find Azure Marketplace pricing when deploying Mistral models as a service.

The cost of using the Mistral Large model in Azure AI Studio is based on the number of tokens for both input and output.

The cost is approximately $0.024 per 1000 tokens for output, while the cost is $0.008 per 1000 tokens for input.

Mistral models deployed as a service track costs associated with their consumption through a new resource created each time a workspace subscribes to a given model offering from Azure Marketplace.

You can review the pricing on the Mistral Large offer in the Marketplace offer details tab when deploying the model or on the Azure Marketplace.

To track costs, see Monitor costs for models offered through the Azure Marketplace.

Each deployment has a rate limit of 200,000 tokens per minute and 1,000 API requests per minute.

Azure Mistral Models provide a way to consume models as an API without hosting them on your subscription, while keeping enterprise security and compliance. This deployment option doesn't require quota from your subscription.

You can deploy Mistral family of models as a serverless API with pay-as-you-go billing through the Microsoft Azure Marketplace. The models that can be deployed are Mistral Large (2402), Mistral Large (2407), Mistral Small, and Mistral Nemo.

The Mistral family of models can be consumed by using the chat API, which allows you to make an API request using the Azure AI Model Inference API on the route /chat/completions and the native Mistral Chat API on /v1/chat/completions.

Mistral models accept both the Azure AI Model Inference API on the route /chat/completions and the native Mistral Chat API on /v1/chat/completions.

The Mistral Large model, Mistral AI's flagship commercial model, is now available first on Azure AI and the Mistral AI platform. This marks a noteworthy expansion of their offerings.

Here are the top-tier reasoning models available under the Mistral Research License and Commercial License:

You can consume Mistral models by using the chat API, which allows you to make API requests to either the Azure AI Model Inference API or the native Mistral Chat API.

The Mistral Large model, Mistral AI's flagship commercial model, is a general-purpose language model that can deliver on any text-based use case thanks to its state-of-the-art reasoning and knowledge capabilities.

To deploy the Mistral Large model, you can follow these steps:

You can have only one deployment for each model within a project, and each project has its own subscription to the particular Azure Marketplace offering of the model, which allows you to control and monitor spending.

Mistral's human portion has water green skin, long black hair, and golden eyes.

Her snake body is blue and orange, and she wraps it around a stone pillar that is capable of summoning fiends.

Mistral wears a very revealing thin white veil with golden accents, and on her body, she has four chakrams and a blue fan that she uses as weapons.

A blue and gold hat with two green and red feathers respectively, and an ornament resembling a scale, adorns her head.

Silver and golden jewelry is worn around her waist and throughout the entirety of her snake body.