Microsoft Foundry

Microsoft Agent Framework adds the CodeAct pattern in public preview, which lets agents collapse a multi-step plan into a single executable code block rather than chaining one tool call at a time. In…

Microsoft Agent Framework adds DevUI Agent Inspector in public preview, a local debugging surface that shows agent runs as a navigable tree of model calls, tool invocations, and intermediate state.…

Microsoft Agent Framework adds episodic procedural memory in public preview, a built-in store that lets agents remember successful task patterns and replay them on similar future requests. The…

In public preview, Microsoft Foundry now offers tracing for the Microsoft Agent Framework in both Python and .NET, helping developers observe how agents reason, call tools, and interact with models…

Microsoft Foundry is now introducing Agent Framework, a unified multi-agent orchestration SDK for C# and Python. This release consolidates patterns previously spread across AutoGen and Semantic…

Microsoft Agent Framework reaches general availability for multi-agent orchestration patterns, including the dynamic Magentic pattern that delivered 38 percent on GAIA and competitive results on…

Microsoft Agent Framework introduces Agent Channel in public preview, a built-in primitive that lets agents send structured messages to each other without bespoke routing code. Channels carry typed…

Microsoft Agent Framework reaches general availability for Agent Harness, the production runtime for agents built with the framework. Customers running pilots on the preview harness can now move the…

Microsoft Agent Framework now ships generally available connectors for GitHub Copilot and Claude Code, so developers can build .NET and Python agents that delegate coding tasks to either system…

Document-based PII detection in Azure AI Language now ships an initial playground experience in the Microsoft Foundry NextGen portal. The playground loads a prepared sample document, runs it through…

Azure AI Language now offers a refreshed Document PII playground in general availability, giving customers a faster way to evaluate personally identifiable information detection on documents. The…

Microsoft Foundry brings Toolboxes to GA, providing a unified endpoint to invoke the most-used tools in prompt agents and callable outside of prompt agents. Today every agent team hand-assembles its…

In public preview, Microsoft Foundry now supports avatar voice sync with custom voices in the Voice Live API, allowing developers to pair branded or persona-specific text-to-speech voices with…

Microsoft Foundry adds tool search to Foundry toolboxes in public preview, giving developers and admins a fast way to find the right tool inside large, multi-team catalogs. As your toolbox grows,…

Microsoft Foundry adds Azure SQL Database as a first-class Knowledge Source in Azure AI Search, available in public preview. Enterprise developers and solution architects building Copilot, RAG, and…

Microsoft Foundry IQ now treats a Microsoft Fabric Ontology as a federated knowledge source in public preview, giving agents a way to query the semantic layer customers already curate in Fabric.…

Microsoft Foundry IQ reaches general availability, giving developers a managed knowledge layer that grounds agents in enterprise data without rebuilding retrieval pipelines for each project. Teams…

Azure AI Search and Foundry Knowledge Bases now support private, end-to-end network connectivity between search resources and Foundry services. Customers can route ingestion, enrichment, retrieval,…

Microsoft Foundry brings Azure API Management support to all Foundry model integrations in Azure AI Search, now in public preview. Enterprise platform engineers and AI solution teams running…

Azure AI Search now supports Content Understanding chunking and image verbalization in public preview, expanding the built-in skill that parses rich documents during indexing. Indexers can split…

Azure AI Language brings the next-generation Text Analytics for Health (TA4H) playground to general availability in the Microsoft Foundry portal. Clinical informatics teams and ISVs can paste…

Microsoft Foundry Agent Service adds agent-to-agent (A2A) communication in public preview for Prompt agents and Hosted agents that use the responses protocol. Agents can call other agents by name…

The Microsoft Foundry extension for Visual Studio Code reaches general availability with a Build 2026 refresh that brings the full model catalog, model playground, and Hosted agent deployment into…

Azure AI Speech moves the Custom Voice authoring experience into the Microsoft Foundry portal in general availability. Approved Custom Voice customers can now upload voice talent recordings and…

Microsoft Foundry brings self-serve custom photo avatar creation to GA in the Foundry NextGen fine-tuning portal. Enterprise customers building voice agents with avatars or producing content with…

In public preview, Microsoft Foundry now offers an updated TextPII NextGen Playground with a refreshed API Configuration Panel, including preview features introduced at Ignite 2025. This update lets…

Microsoft Foundry introduces rubric evaluators for measuring agent quality, now in public preview for single-turn and multi-turn agent flows. Rubric evaluators enable developers to define…

Microsoft Foundry now offers MAI-Voice-2 in public preview, a first-party voice model from the Microsoft AI team. MAI-Voice-2 generates natural speech across more than 10 languages and supports voice…

Microsoft Foundry model catalog adds a domain filter in public preview, helping developers narrow the catalog of over 1,900 models to the ones trained for a specific industry or use case. Filters…

In public preview, Microsoft Foundry now offers a Conversational PII (ConvPII) NextGen Playground with transcript input and an API Configuration Panel, enabling developers to validate PII detection…

Microsoft Foundry observability introduces intelligent trace filtering and sampling for evaluations in public preview. Instead of running evaluations against all production traces, Foundry now select…

Use the Azure Developer CLI (azd) CLI evaluation experience to add a measured quality loop to an agent created with Microsoft Foundry. It empowers users on the hosted-agent lifecycle in azd, where…

This feature introduces end user feedback logging semantics that enables developers to capture structured feedback signals from real users interacting with AI agents—such as thumbs up/down, ratings,…

Azure AI Search makes the GenAI prompt skill generally available, letting indexers call chat completion models during enrichment without custom code. Authors point the skill at a Foundry-hosted…

Azure AI Search now supports managed identity authentication for the billing operations it issues against Microsoft Foundry resources, replacing the prior key-based flow. Customers assign a…

Microsoft Foundry adds region-agnostic reservations for Global Provisioned Throughput Units (PTUs) at general availability, letting customers commit to PTU capacity once and consume it from any…

Global Provisioned Throughput (PTU) reservations are now region-agnostic, enabling a single reservation to cover Global PTU deployments across multiple regions. Why it matters:Optimize savings: One…

Azure AI Translator's unified Text Translation API reaches general availability, giving developers a single endpoint that fronts neural machine translation, large language model translation, small…

Azure AI Translator batch document translation ships generally available improvements for PDF translation. The service uses Azure AI Document Intelligence to recover structure from both…

Foundry IQ knowledge bases now surface document-extracted images at query time in public preview. When ingestion pulls figures, screenshots, and diagrams out of PDFs, slides, and Office files,…

Azure Policy now provides centralized governance for deployments using the model router. Organizations can define and enforce routing standards across their environment, helping ensure that model…

Azure AI Speech makes the LLM Speech API generally available, delivering state-of-the-art LLM-powered transcription and translation for audio files, with multi-lingual support across 25 languages…

Microsoft Foundry Agent Service is now generally available for Voice Live integration, allowing developers to connect real-time speech-to-text and text-to-speech to a Foundry Agent without writing…

Microsoft Foundry brings code-first observability to agent developers in public preview, delivered as an observe skill inside the Foundry Plugin for VS Code Copilot Chat. Developers run the full…

Azure AI Translator batch document translation accepts image files as a generally available input type. Supported formats include `.jpeg`, `.png`, `.bmp`, and `.webp`. The service runs OCR on each…

Microsoft Foundry refreshes the Foundry Memory public preview with a redesigned storage model, new retrieval controls, and tighter integration with the Foundry Agent Service tool surface. Agents…

Microsoft Foundry observability adds Trace Replay and new trace visualizations in public preview, giving agent developers a faster way to debug multi-step runs. The replay view rewinds a captured…

Content Understanding in Microsoft Foundry now supports an extractionMode setting that lets you choose how documents are analyzed. Pick a layout-aware mode for structured documents where tables,…

Microsoft Foundry makes managed virtual network support generally available for evaluations, so customers can run cloud evaluation jobs against models and agents that sit behind private endpoints.…

Azure AI Search now offers a serverless option for native indexers in public preview, removing the requirement to provision and manage indexer compute for ingestion workflows. Indexer execution…

Adaptive Custom Translation (AdaptCT) is generally available in the Azure AI Foundry NextGen playground for Azure Translator. AdaptCT enables enterprises to adapt translations to domain‑specific…

Speech SDK 1.50 ships as a generally available release across all supported languages: C#, C++, Java, JavaScript, Python, Objective-C, Swift, and Go. The release rolls up the latest speech-to-text…

Azure AI Search makes OneLake catalog integration generally available for knowledge sources, so customers can register a OneLake item once and reuse it across multiple knowledge sources and agents.…

Microsoft Foundry Agent Service reaches preview for Hosted Agents, the container-based runtime that lets developers bring their own agent code in any framework and deploy it to Microsoft-managed…

Azure AI Search and Foundry IQ Knowledge Sources add incremental SharePoint permissions sync in public preview, keeping document-level, site, and library access control lists current as SharePoint…

Azure AI Translator batch document translation now translates text inside images embedded in Word `.docx` files as a generally available capability. The service detects image regions, runs OCR,…

Azure AI Search adds Microsoft Purview admin access auditing for sensitivity-labeled content in public preview. Admin operations that touch Purview-labeled documents inside Azure AI Search, knowledge…

Microsoft Foundry IQ now supports external MCP servers as federated knowledge sources in public preview, expanding retrieval beyond data customers index themselves. Teams can register MCP servers…

Microsoft Foundry adds connectors and time-based triggers to the Foundry Tools toolbox in public preview, giving enterprise developers and automation engineers a native way to build event-driven AI…

Azure AI Search now emits audit events for Microsoft Purview sensitivity labels carried alongside indexed documents. When the SharePoint in Microsoft 365 indexer is configured to preserve and honor…

Azure AI Speech brings the Custom Avatar and Custom Video authoring experiences into the Microsoft Foundry portal in general availability. Approved customers can upload an actor's video recording,…

The Azure AI Search SharePoint in Microsoft 365 indexer now indexes ASPX site content and SharePoint Lists in addition to document libraries. The indexer connects through a Microsoft Entra…

Choosing the right AI system is rarely as simple as picking the newest or largest model — and agents add their own layer of variability on top. A model that wins on a public leaderboard can still…

Azure AI Search adds Microsoft Purview sensitivity label support to knowledge sources in public preview. Labels applied at the source system flow through ingestion into Azure AI Search and into the…

Microsoft Foundry adds MAI-Transcribe-1.5 to the model catalog in public preview, Microsoft's next-generation speech-to-text model. The release adds improved accuracy with noticeably lower word error…

Azure AI Translator now supports synchronous single-document image translation at general availability. Customers can submit a standalone image file (JPEG, PNG, and similar formats) through one…

Microsoft Foundry IQ adds public preview support for OneLake Catalog as a knowledge source, so agents can ground responses on lakehouse tables, files, and shortcuts already governed in Microsoft…

Microsoft Foundry adds MAI-Image-2.5 to the model catalog in public preview, the next generation of Microsoft's in-house image generation model. The release improves photorealism, prompt fidelity,…

Azure AI Translator publishes generally available SDKs that target the latest text translation REST API across four languages: C#/.NET, Java, JavaScript, and Python. The SDKs expose Translator API v4…

Azure AI Speech is rolling out an updated Neural HD text-to-speech model, moving customers from DragonHDLatest to HDv2.5. The update raises audio quality scores, adds style and paralinguistic tag…

Microsoft Foundry brings Content Understanding into the portal in public preview, ending the context switch to the standalone Content Understanding Studio for everyday work. Developers get Read and…

Photo Avatar in Azure AI Speech is now generally available in standard and custom flavors. Standard Photo Avatar produces a talking head video from a single photo at 512x512 resolution, and custom…

Microsoft Agent Framework is now version 1.0 for both .NET and Python, with stable APIs and a long-term support commitment. Agent Framework 1.0 supports multi-agent orchestration, multi-provider…

Foundry’s memory (preview) is a managed long-term memory capability built directly into Foundry Agent Service and now integrates natively with Microsoft Agent Framework and LangGraph— no external…

Foundry Agent Service now supports bring‑your‑own‑model scenarios, allowing agents to use AI models deployed and managed outside of Foundry, including models hosted behind enterprise AI gateways such…

Hosted agents run in an isolated execution sandbox that provides a dedicated, secure runtime per agent session. Sessions start with a clean environment with no shared state between sessions, no…

Foundry Toolkit for VS Code is now generally available and provides VS Code tooling for Microsoft Agent Framework, including agent creation from templates or via GitHub Copilot, local test/debug with…

Announcing general availability (GA) of the next-gen Foundry Agent Service—a redesigned API format and runtime experience designed to help teams build and operate agents that can move from prototype…

Claude Opus 4.6 is now available in Microsoft Foundry, bringing Anthropic’s most advanced reasoning model to a secure, enterprise-ready platform built on Azure. With support for complex coding,…