Azure OpenAI Provider

Learn how to use Azure OpenAI Service with the AI SDK for enterprise deployments.

Azure OpenAI Provider

The Azure OpenAI provider enables you to use OpenAI models through Microsoft Azure with enterprise features, security, and compliance.

Installation

npm

npm install @ai-sdk/azure

Setup

Get your credentials from the Azure Portal and set them as environment variables:

AZURE_RESOURCE_NAME=your-resource-name
AZURE_API_KEY=your-api-key

Usage

Basic Text Generation

import { azure } from '@ai-sdk/azure';
import { generateText } from 'ai';

const { text } = await generateText({
  model: azure('your-gpt-4-deployment'),
  prompt: 'Explain enterprise AI solutions.',
});

Streaming Responses

import { azure } from '@ai-sdk/azure';
import { streamText } from 'ai';

const result = streamText({
  model: azure('your-gpt-4-deployment'),
  prompt: 'Write about cloud security.',
});

for await (const chunk of result.textStream) {
  process.stdout.write(chunk);
}

Structured Output

import { azure } from '@ai-sdk/azure';
import { generateObject } from 'ai';
import { z } from 'zod';

const { object } = await generateObject({
  model: azure('your-gpt-4-deployment'),
  schema: z.object({
    risks: z.array(z.string()),
    recommendations: z.array(z.string()),
  }),
  prompt: 'Analyze security risks.',
});

Configuration

Custom Provider Instance

import { createAzure } from '@ai-sdk/azure';

const azure = createAzure({
  resourceName: 'your-resource-name',
  apiKey: process.env.AZURE_API_KEY,
});

Provider Options

• resourceName: Your Azure resource name
• apiKey: Azure OpenAI API key
• apiVersion: API version (defaults to ‘v1’)
• baseURL: Custom endpoint URL

Available Models

You can deploy the following OpenAI models on Azure:

Chat Models

• GPT-4 Turbo
• GPT-4
• GPT-3.5 Turbo

Embeddings

import { azure } from '@ai-sdk/azure';
import { embed } from 'ai';

const { embedding } = await embed({
  model: azure.embedding('your-embedding-deployment'),
  value: 'Enterprise document',
});

Available Models:

• text-embedding-3-small
• text-embedding-3-large
• text-embedding-ada-002

Image Generation

import { azure } from '@ai-sdk/azure';
import { generateImage } from 'ai';

const { image } = await generateImage({
  model: azure.image('your-dalle-deployment'),
  prompt: 'Professional office space',
  size: '1024x1024',
});

Available Models:

• DALL-E 3
• DALL-E 2

Advanced Features

Tool Calling

import { azure } from '@ai-sdk/azure';
import { generateText } from 'ai';
import { z } from 'zod';

const result = await generateText({
  model: azure('your-gpt-4-deployment'),
  tools: {
    database_query: {
      description: 'Query the database',
      parameters: z.object({
        query: z.string(),
      }),
      execute: async ({ query }) => {
        return { results: [] };
      },
    },
  },
  prompt: 'Find all users from last month.',
});

Vision

import { azure } from '@ai-sdk/azure';
import { generateText } from 'ai';
import fs from 'fs';

const result = await generateText({
  model: azure('your-gpt-4-vision-deployment'),
  messages: [
    {
      role: 'user',
      content: [
        { type: 'text', text: 'Analyze this chart.' },
        { type: 'image', image: fs.readFileSync('./chart.png') },
      ],
    },
  ],
});

PDF Support

import { azure } from '@ai-sdk/azure';
import { generateText } from 'ai';
import fs from 'fs';

const result = await generateText({
  model: azure('your-gpt-4-deployment'),
  messages: [
    {
      role: 'user',
      content: [
        { type: 'text', text: 'Review this contract.' },
        {
          type: 'file',
          data: fs.readFileSync('./contract.pdf'),
          mediaType: 'application/pdf',
        },
      ],
    },
  ],
});

Deployment Notes

• Create deployments in the Azure Portal before use
• Each deployment has its own name and configuration
• Deployment names are used as model identifiers
• Supports virtual network integration
• Provides managed identity authentication

Resources

• Azure OpenAI Documentation
• AI SDK Documentation
• Azure Portal