How to beginner · 3 min read

How to use Azure OpenAI embeddings

Q: How to use Azure OpenAI embeddings

Use the AzureOpenAI client from the openai package with your Azure endpoint and API key. Call client.embeddings.create() specifying your deployment name and input text to get embeddings from Azure OpenAI.

Quick answer

Use the AzureOpenAI client from the openai package with your Azure endpoint and API key. Call client.embeddings.create() specifying your deployment name and input text to get embeddings from Azure OpenAI.

PREREQUISITES

Python 3.8+
Azure OpenAI resource with deployment for embeddings
Azure OpenAI API key and endpoint
pip install openai>=1.0

Setup

Install the openai Python package and set environment variables for your Azure OpenAI API key and endpoint. You also need the deployment name of your embeddings model.

bash

pip install openai>=1.0

Step by step

This example shows how to create embeddings using Azure OpenAI. Replace environment variables with your Azure OpenAI details and specify your embedding deployment name.

python

import os
from openai import AzureOpenAI

client = AzureOpenAI(
    api_key=os.environ["AZURE_OPENAI_API_KEY"],
    azure_endpoint=os.environ["AZURE_OPENAI_ENDPOINT"],
    api_version="2024-02-01"
)

response = client.embeddings.create(
    model=os.environ["AZURE_OPENAI_EMBEDDING_DEPLOYMENT"],
    input="The quick brown fox jumps over the lazy dog"
)

embedding_vector = response.data[0].embedding
print("Embedding vector length:", len(embedding_vector))
print("First 5 values:", embedding_vector[:5])

output

Embedding vector length: 1536
First 5 values: [0.0123, -0.0345, 0.0567, 0.0789, -0.0123]

Common variations

Use different embedding models by changing the deployment name in model.
Batch multiple inputs by passing a list to input.
Use async calls with asyncio and await if your environment supports it.

python

import asyncio
import os
from openai import AzureOpenAI

async def create_embeddings():
    client = AzureOpenAI(
        api_key=os.environ["AZURE_OPENAI_API_KEY"],
        azure_endpoint=os.environ["AZURE_OPENAI_ENDPOINT"],
        api_version="2024-02-01"
    )

    response = await client.embeddings.acreate(
        model=os.environ["AZURE_OPENAI_EMBEDDING_DEPLOYMENT"],
        input=["Hello world", "Azure OpenAI embeddings"]
    )

    for i, embedding in enumerate(response.data):
        print(f"Embedding {i} length:", len(embedding.embedding))

asyncio.run(create_embeddings())

output

Embedding 0 length: 1536
Embedding 1 length: 1536

Troubleshooting

If you get authentication errors, verify your AZURE_OPENAI_API_KEY and AZURE_OPENAI_ENDPOINT environment variables.
Ensure your deployment name matches exactly the one configured in the Azure portal.
Check your API version matches the SDK usage (2024-02-01 recommended).

✅

Key Takeaways

Use the AzureOpenAI client with your Azure endpoint and API key to access embeddings.
Specify your embedding deployment name as the model parameter in embeddings.create().
Batch inputs by passing a list to input for efficient embedding generation.
Use the api_version="2024-02-01" parameter to ensure compatibility with Azure OpenAI features.
Verify environment variables and deployment names carefully to avoid authentication and usage errors.

Verified 2026-04 · gpt-4o, text-embedding-3-large

Verify ↗