News
What's New in the Azure SDK
Microsoft released the October 2025 update of the Azure SDK, delivering new stable libraries, expanded language coverage, and performance improvements across the platform.
The update continues the monthly cadence of Azure SDK enhancements, bringing developers new capabilities in AI Foundry, AI Search, Identity, and resource management libraries for multiple languages.
A highlight of this release is the general availability of the Azure AI Foundry client library for .NET, version 1.0.0. With this new library, developers can build, deploy, and manage AI agents within their .NET applications. It supports managing connections to Azure resources, handling model deployments, and uploading documents to create datasets and indexes. Key updates include the introduction of direct properties on the AIProjectClient for easier use and the consolidation of index management functions under CreateOrUpdate methods. These changes simplify code and improve consistency when building AI agent workflows within Azure environments.
Azure AI Search also received major updates across several programming languages. The JavaScript library, version 12.2.0, now supports service version 2025-09-01 and introduces vector query support for subfields of complex fields. It adds reranker-boosted scoring, vector compression configuration, and improved lexical normalization options. The Python library, version 11.6.0, includes equivalent features along with a new DocumentIntelligenceLayoutSkill and improved integration for structured document parsing. These enhancements help developers perform advanced semantic search tasks with improved accuracy and scalability across large datasets.
Significant work also went into improving performance and efficiency in Azure Identity libraries. The October release optimizes managed identity handling across all major languages including C++, Go, Java, JavaScript, and Python. When using managed identities, the DefaultAzureCredential can now skip the Instance Metadata Service (IMDS) probe when the environment variable AZURE_TOKEN_CREDENTIALS is set to ManagedIdentityCredential. This allows faster token acquisition with built-in retry logic and exponential backoff, reducing latency for applications that rely on secure token-based authentication. The change improves consistency and reliability for developers working with managed identities in production workloads.
The Azure SDK team also finalized several stable releases for resource management libraries. These include new versions across .NET, Java, JavaScript, Python, Go, and C++. Among them are new general availability packages for durable task management, storage discovery, and additional service management APIs. Each library is designed to align with the latest Azure service versions and improve compatibility for automation and deployment scenarios.
Microsoft said the October release continues the Azure SDK team's focus on performance, reliability, and usability. The release also includes minor bug fixes, API refinements, and additional support for new Azure services. Developers are encouraged to review the updated changelogs and documentation before upgrading, as some libraries include structural adjustments that may affect existing implementations.
Teams that use Azure AI Foundry in .NET applications are advised to update to version 1.0.0 to take advantage of the final stable APIs. Developers using Azure AI Search can adopt the latest versions to leverage vector search enhancements and new configuration options. Those relying on Azure Identity libraries may benefit from reduced latency and improved reliability for managed identity authentication.
The Azure SDK team maintains a regular monthly release cycle, providing iterative updates that support both new service development and long-term platform stability.
About the Author
David Ramel is an editor and writer at Converge 360.