Skip to main content

​
Overview

Azure OpenAI Service provides access to OpenAI’s models through Microsoft Azure infrastructure. Zerox supports Azure OpenAI deployments for document processing with enhanced security and compliance features.

​
Credentials

Azure OpenAI requires both an API key and an endpoint URL:
​
credentials.apiKey
string
required
Your Azure OpenAI API key. Found in the Azure Portal under your OpenAI resource’s “Keys and Endpoint” section.
​
credentials.endpoint
string
required
Your Azure OpenAI endpoint URL. Format: https://your-resource-name.openai.azure.com

​
Environment Variables

export AZURE_API_KEY="your-azure-api-key"
export AZURE_ENDPOINT="https://your-resource-name.openai.azure.com"

​
Supported Models

Azure OpenAI supports the same vision models as OpenAI, but they must be deployed in your Azure resource:
ModelDeployment NameDescription
GPT-4.1gpt-4.1Latest GPT-4.1 model with vision capabilities
GPT-4.1 Minigpt-4.1-miniSmaller, faster GPT-4.1 model
GPT-4ogpt-4oOptimized GPT-4 model with vision
GPT-4o Minigpt-4o-miniSmaller, cost-effective GPT-4o model
The model parameter in Zerox should match your Azure deployment name, not the base model name. For example, if you deployed gpt-4o with the name my-gpt4o-deployment, use that deployment name.

​
Configuration

​
Basic Example

import { zerox } from "zerox";
import { ModelOptions, ModelProvider } from "zerox/node-zerox/dist/types";

const result = await zerox({
  filePath: "path/to/document.pdf",
  modelProvider: ModelProvider.AZURE,
  model: ModelOptions.OPENAI_GPT_4O,
  credentials: {
    apiKey: process.env.AZURE_API_KEY,
    endpoint: process.env.AZURE_ENDPOINT,
  },
});

​
With Custom Deployment Name

import { zerox } from "zerox";
import { ModelProvider } from "zerox/node-zerox/dist/types";

const result = await zerox({
  filePath: "path/to/document.pdf",
  modelProvider: ModelProvider.AZURE,
  model: "my-gpt4o-deployment", // Your Azure deployment name
  credentials: {
    apiKey: process.env.AZURE_API_KEY,
    endpoint: process.env.AZURE_ENDPOINT,
  },
});

​
LLM Parameters

Azure OpenAI models support the following optional parameters:
​
llmParams.temperature
number
default:"1"
Controls randomness in the output. Values range from 0 to 2. Lower values make output more focused and deterministic.
​
llmParams.maxTokens
number
default:"4096"
Maximum number of tokens to generate in the completion.
​
llmParams.topP
number
default:"1"
Nucleus sampling parameter. An alternative to temperature sampling. Values range from 0 to 1.
​
llmParams.frequencyPenalty
number
default:"0"
Number between -2.0 and 2.0. Positive values penalize tokens based on their frequency in the text so far.
​
llmParams.presencePenalty
number
default:"0"
Number between -2.0 and 2.0. Positive values penalize tokens based on whether they appear in the text so far.
​
llmParams.logprobs
boolean
default:"false"
Whether to return log probabilities of the output tokens. Useful for confidence scoring.

​
Example with Parameters

import { zerox } from "zerox";
import { ModelOptions, ModelProvider } from "zerox/node-zerox/dist/types";

const result = await zerox({
  filePath: "path/to/document.pdf",
  modelProvider: ModelProvider.AZURE,
  model: ModelOptions.OPENAI_GPT_4O_MINI,
  credentials: {
    apiKey: process.env.AZURE_API_KEY,
    endpoint: process.env.AZURE_ENDPOINT,
  },
  llmParams: {
    temperature: 0.3,
    maxTokens: 4096,
    topP: 0.9,
    logprobs: true,
  },
});

​
Data Extraction

Azure OpenAI models support structured data extraction using JSON schemas: 
import { zerox } from "zerox";
import { ModelOptions, ModelProvider } from "zerox/node-zerox/dist/types";

const schema = {
  type: "object",
  properties: {
    company_name: { type: "string" },
    registration_number: { type: "string" },
    address: { type: "string" },
    officers: {
      type: "array",
      items: {
        type: "object",
        properties: {
          name: { type: "string" },
          title: { type: "string" },
        },
      },
    },
  },
};

const result = await zerox({
  filePath: "company-filing.pdf",
  modelProvider: ModelProvider.AZURE,
  model: ModelOptions.OPENAI_GPT_4O,
  credentials: {
    apiKey: process.env.AZURE_API_KEY,
    endpoint: process.env.AZURE_ENDPOINT,
  },
  extractOnly: true,
  schema: schema,
});

console.log(result.extracted);

​
API Version

Zerox uses Azure OpenAI API version 2024-10-21 by default. This is automatically configured and does not need to be specified.

​
Error Handling

import { zerox } from "zerox";
import { ErrorMode, ModelProvider } from "zerox/node-zerox/dist/types";

try {
  const result = await zerox({
    filePath: "document.pdf",
    modelProvider: ModelProvider.AZURE,
    model: "gpt-4o",
    credentials: {
      apiKey: process.env.AZURE_API_KEY,
      endpoint: process.env.AZURE_ENDPOINT,
    },
    errorMode: ErrorMode.THROW,
  });
} catch (error) {
  console.error("OCR failed:", error);
}
Before using Azure OpenAI with Zerox:
  • Ensure you have deployed a vision-capable model in your Azure OpenAI resource
  • Verify your deployment name matches the model parameter
  • Confirm your resource has sufficient quota for vision API requests
  • Check that your endpoint URL is correct and accessible

​
Azure-Specific Considerations

​
Deployment Requirements

You must create a deployment in your Azure OpenAI resource before using it with Zerox:
  1. Navigate to Azure OpenAI Studio
  2. Create a new deployment
  3. Select a vision-capable model (GPT-4o, GPT-4o-mini, etc.)
  4. Use the deployment name (not the model name) in Zerox configuration

​
Regional Availability

Vision models may not be available in all Azure regions. Check the Azure OpenAI Service regions documentation for current availability.

​
Quota Management

Azure OpenAI uses tokens-per-minute (TPM) quotas. Monitor your usage to avoid rate limiting: 
const result = await zerox({
  filePath: "document.pdf",
  modelProvider: ModelProvider.AZURE,
  model: "gpt-4o",
  credentials: {
    apiKey: process.env.AZURE_API_KEY,
    endpoint: process.env.AZURE_ENDPOINT,
  },
  concurrency: 5, // Limit concurrent requests
});

console.log(`Tokens used: ${result.inputTokens + result.outputTokens}`);

​
Best Practices

  • Use managed identities instead of API keys for production deployments
  • Monitor your Azure costs through the Azure Portal
  • Set appropriate concurrency values to stay within quota limits
  • Use gpt-4o-mini deployments for cost optimization
  • Enable diagnostic logging in Azure for troubleshooting

Build docs developers (and LLMs) love

Get started for freeTalk to us