Azure Identity client library for Python - version 1.25.2

The Azure Identity library provides Microsoft Entra ID token-based authentication support across the Azure SDK. It provides a set of TokenCredential/SupportsTokenInfo implementations, which can be used to construct Azure SDK clients that support Microsoft Entra token authentication.

Source code | Package (PyPI) | Package (Conda) | API reference documentation | Microsoft Entra ID documentation

Getting started

Install the package

Install Azure Identity with pip:

pip install azure-identity

Prerequisites

• An Azure subscription
• Python 3.9 or a recent version of Python 3 (this library doesn't support end-of-life versions)

Authenticate the client

When debugging and executing code locally, it's typical for a developer to use their own account for authenticating calls to Azure services. There are several developer tools that can be used to perform this authentication in your development environment. For more information, see Authentication during local development.

Key concepts

Credentials

A credential is a class that contains or can obtain the data needed for a service client to authenticate requests. Service clients across the Azure SDK accept a credential instance when they're constructed, and use that credential to authenticate requests.

The Azure Identity library focuses on OAuth authentication with Microsoft Entra ID. It offers various credential classes capable of acquiring a Microsoft Entra access token. See the Credential classes section for a list of this library's credential classes.

DefaultAzureCredential

DefaultAzureCredential simplifies authentication while developing apps that deploy to Azure by combining credentials used in Azure hosting environments with credentials used in local development. For more information, see DefaultAzureCredential overview.

Continuation policy

As of version 1.14.0, DefaultAzureCredential attempts to authenticate with all developer credentials until one succeeds, regardless of any errors previous developer credentials experienced. For example, a developer credential may attempt to get a token and fail, so DefaultAzureCredential will continue to the next credential in the flow. Deployed service credentials stop the flow with a thrown exception if they're able to attempt token retrieval, but don't receive one. Prior to version 1.14.0, developer credentials would similarly stop the authentication flow if token retrieval failed, but this is no longer the case.

This allows for trying all of the developer credentials on your machine while having predictable deployed behavior.

Examples

The following examples are provided:

• Define a custom authentication flow with ChainedTokenCredential
• Async credentials

Define a custom authentication flow with ChainedTokenCredential

While DefaultAzureCredential is generally the quickest way to authenticate apps for Azure, you can create a customized chain of credentials to be considered. ChainedTokenCredential enables users to combine multiple credential instances to define a customized chain of credentials. For more information, see ChainedTokenCredential overview.

Async credentials

This library includes a set of async APIs. To use the async credentials in azure.identity.aio, you must first install an async transport, such as aiohttp. For more information, see azure-core documentation.

Async credentials should be closed when they're no longer needed. Each async credential is an async context manager and defines an async close method. For example:

from azure.identity.aio import DefaultAzureCredential

# call close when the credential is no longer needed
credential = DefaultAzureCredential()
...
await credential.close()

# alternatively, use the credential as an async context manager
credential = DefaultAzureCredential()
async with credential:
  ...

This example demonstrates authenticating the asynchronous SecretClient from azure-keyvault-secrets with an asynchronous credential.

from azure.identity.aio import DefaultAzureCredential
from azure.keyvault.secrets.aio import SecretClient

default_credential = DefaultAzureCredential()
client = SecretClient("https://my-vault.vault.azure.net", default_credential)

Managed identity support

Managed identity authentication is supported either indirectly via DefaultAzureCredential or directly via ManagedIdentityCredential for the following Azure services:

• Azure App Service and Azure Functions
• Azure Arc
• Azure Cloud Shell
• Azure Kubernetes Service
• Azure Service Fabric
• Azure Virtual Machines
• Azure Virtual Machines Scale Sets

Cloud configuration

Credentials default to authenticating to the Microsoft Entra endpoint for Azure Public Cloud. To access resources in other clouds, such as Azure Government or a private cloud, configure credentials with the authority argument. AzureAuthorityHosts defines authorities for well-known clouds:

from azure.identity import AzureAuthorityHosts

DefaultAzureCredential(authority=AzureAuthorityHosts.AZURE_GOVERNMENT)

If the authority for your cloud isn't listed in AzureAuthorityHosts, you can explicitly specify its URL:

DefaultAzureCredential(authority="https://login.partner.microsoftonline.cn")

As an alternative to specifying the authority argument, you can also set the AZURE_AUTHORITY_HOST environment variable to the URL of your cloud's authority. This approach is useful when configuring multiple credentials to authenticate to the same cloud:

AZURE_AUTHORITY_HOST=https://login.partner.microsoftonline.cn

Not all credentials require this configuration. Credentials that authenticate through a development tool, such as AzureCliCredential, use that tool's configuration.

Credential classes

Credential chains

Authenticate Azure-hosted applications

Authenticate service principals

Authenticate users

Authenticate via development tools

Environment variables

DefaultAzureCredential and EnvironmentCredential can be configured with environment variables. Each type of authentication requires values for specific variables:

Service principal with secret

Service principal with certificate

Configuration is attempted in the preceding order. For example, if values for a client secret and certificate are both present, the client secret is used.

Continuous Access Evaluation

As of version 1.14.0, accessing resources protected by Continuous Access Evaluation (CAE) is possible on a per-request basis. This behavior can be enabled by setting the enable_cae keyword argument to True in the credential's get_token method. CAE isn't supported for developer and managed identity credentials.

Token caching

Token caching is a feature provided by the Azure Identity library that allows apps to:

• Cache tokens in memory (default) or on disk (opt-in).
• Improve resilience and performance.
• Reduce the number of requests made to Microsoft Entra ID to obtain access tokens.

The Azure Identity library offers both in-memory and persistent disk caching. For more information, see the token caching documentation.

Brokered authentication

An authentication broker is an application that runs on a user’s machine and manages the authentication handshakes and token maintenance for connected accounts. Currently, only the Windows Web Account Manager (WAM) is supported. Authentication via WAM is used by DefaultAzureCredential to enable secure sign-in. To enable support, use the azure-identity-broker package. For details on authenticating using WAM, see the broker plugin documentation.

Troubleshooting

See the troubleshooting guide for details on how to diagnose various failure scenarios.

Error handling

Credentials raise CredentialUnavailableError when they're unable to attempt authentication because they lack required data or state. For example, EnvironmentCredential raises this exception when its configuration is incomplete.

Credentials raise azure.core.exceptions.ClientAuthenticationError when they fail to authenticate. ClientAuthenticationError has a message attribute, which describes why authentication failed. When raised by DefaultAzureCredential or ChainedTokenCredential, the message collects error messages from each credential in the chain.

For more information on handling specific Microsoft Entra ID errors, see the Microsoft Entra ID error code documentation.

Logging

This library uses the standard logging library for logging. Credentials log basic information, including HTTP sessions (URLs, headers, etc.) at INFO level. These log entries don't contain authentication secrets.

Detailed DEBUG-level logging, including request/response bodies and header values, isn't enabled by default. It can be enabled with the logging_enable argument. For example:

credential = DefaultAzureCredential(logging_enable=True)

CAUTION: DEBUG-level logs from credentials contain sensitive information. These logs must be protected to avoid compromising account security.

Next steps

Client library support

Client and management libraries listed on the Azure SDK release page that support Microsoft Entra authentication accept credentials from this library. You can learn more about using these libraries in their documentation, which is linked from the release page.

Known issues

This library doesn't support Azure AD B2C.

For other open issues, refer to the library's GitHub repository.

Provide feedback

If you encounter bugs or have suggestions, open an issue.

Contributing

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.microsoft.com.

When you submit a pull request, a CLA-bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., label, comment). Simply follow the instructions provided by the bot. You'll only need to do this once across all repos using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information, see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.