What's new in Microsoft Fabric?

This article is continuously updated with preview/general availability status of Microsoft Fabric.

New to Microsoft Fabric?

Features currently in preview

The following table lists the features of Microsoft Fabric that are currently in preview. Preview features are sorted alphabetically.

Note

Features currently in preview are available under supplemental terms of use. Review the legal terms that apply to Azure features that are in beta, preview, or otherwise not yet released into general availability. Microsoft Fabric provides previews to give you a chance to evaluate and share feedback with the product group on preview features before they become generally available (GA).

Feature Learn more
Activator as business events publisher (Preview) Activator provides a no-code way to publish business events in Microsoft Fabric. When Activator detects that a condition is met in your data, it can emit a structured business event into Real-Time hub, making that signal discoverable, routable, and consumable by your entire organization. For more information, see What is Fabric Activator? and Business events overview.
Activator rule management (Preview) Activator rule management (Preview) provides a central Real-Time Hub view for rules created across Fabric experiences.
Activator rules from catalog and workspace (Preview) Activator rules from catalog and workspace views (Preview) let you create and manage rules from the context of the data they monitor. For more information, see Create and activate an Activator rule.
Advanced DAX generation for semantic models (Preview) Advanced DAX generation for Fabric data agents (Preview) lets a data agent plan its approach, inspect semantic model values and intermediate results, refine its reasoning, and execute a final DAX query before answering. For more information, see Fabric data agent runtime and Semantic model best practices for data agent.
Advanced edit for data destination queries in Dataflow Gen2 (Preview) You can now use advanced edit for data destination queries in Dataflow Gen2 to modify destination-side query logic directly in authoring.
AI-assisted Synapse to Fabric migration skills (Preview) AI-assisted Synapse to Fabric migration skills (Preview) move Synapse Spark pools, lake databases, notebooks, Spark job definitions, and Data Factory pipelines to Microsoft Fabric from the command line by using REST APIs, with lift-and-shift and migrate-and-modernize paths and a read-only assessment step. For more information, see Skills for Microsoft Fabric.
AI Functions multimodal support (Preview) AI Functions apply one-line, LLM-powered transformations to large pandas or PySpark DataFrames with high concurrency by default. For more information, see Use multimodal input with AI Functions (Preview).
ALTER COLUMN for Fabric Data Warehouse (Preview) ALTER TABLE ... ALTER COLUMN (Preview) lets you make supported schema changes directly on warehouse tables with familiar T-SQL syntax. Supported scenarios complete as metadata-only updates without rewriting Parquet files. For more information, see ALTER TABLE (Transact-SQL) syntax for warehouse in Fabric.
ALTER DATABASE SET options in SQL database ALTER DATABASE SET options in SQL database are now available as a preview feature. For more information, see ALTER DATABASE SET options (Transact-SQL) for SQL database in Fabric. Collation support and full-text indexing are also now available in SQL database as preview features.
Anomaly detection (Preview) With a no-code interface, automatic model selection, and flexible alerts, tracking changes and unexpected events is easy with Anomaly detection in Real-Time Intelligence (Preview). For more information, see AI-Powered Real-Time Intelligence with Anomaly Detection (Preview). Billing for Anomaly Detector starts in December.
Anonymous data access for Fabric Apps (Preview) Anonymous data access for Fabric Apps (Preview) lets an app expose selected entity operations to unauthenticated visitors while tenant settings and data-model roles limit access. For more information, see Anonymous data access.
Approval activity in Data Factory pipelines (Preview) The new Approval activity (Preview) in Fabric Data Factory pipelines pauses execution until a reviewer approves or rejects through Outlook or Teams, then routes pipeline flow dynamically based on the outcome (approved, rejected, or timed out). For more information, see Approval activity documentation.
Associate an identity for items (Preview) Associated identities for items (Preview) let you associate a user, service principal, or managed identity with Fabric Lakehouses and Eventstreams via REST API, removing the dependency on the item owner. For more information, see Manage identities associated with Fabric items.
Automatic index compaction in SQL database Automatic index compaction helps you reduce the consumption of storage space, disk I/O, memory, and improve workload performance without investing time and effort into index maintenance jobs.
Autoscaling for Virtual Network Data Gateway (Preview) Autoscaling for Virtual Network Data Gateway (Preview) automatically adjusts the number of active gateway nodes within admin-defined minimum and maximum limits based on workload demand, so the gateway absorbs spikes without manual intervention and scales down to reduce cost during idle periods.
AWS Glue mirroring (Preview) A mirrored AWS Glue catalog brings AWS Glue Iceberg tables into OneLake for Fabric analysis without duplicating data. For more information, see Mirrored AWS Glue catalog.
Azure Event Hubs source workspace identity authentication (Preview) Azure Event Hubs source workspace identity authentication lets an Eventstream Azure Event Hubs source use workspace identity instead of shared access keys for Microsoft Entra ID-based access. For more information, see Add an Azure Event Hubs source to an eventstream.
Azure Monitor mirroring (Preview) Azure Monitor Logs mirroring (Preview) brings Log Analytics workspace tables into OneLake for Fabric analysis without duplicating data. For more information, see Mirror Azure Monitor in Microsoft Fabric.
Azure Monitor to Fabric Eventhouse (Preview) Azure Monitor to Fabric Eventhouse (Preview) lets you route VM telemetry via Azure Monitor Agent and Data Collection Rules into Eventhouse for schema-managed ingestion, ad hoc queries, time-series analytics, and activation. For more information, see Send virtual machine client data to Fabric and Azure Data Explorer (Preview).
Back up Dataflow query results (Preview) Dataflow query-result backup (Preview) preserves prior query results in a lakehouse so you can compare or recover earlier refresh output.
Branch workspace admin profile (Preview) Branch workspace admin profiles (Preview) let administrators define governed identities and settings that developers use to create branch workspaces. For more information, see Development process using branch workspaces.
Build Agent with AI (Preview) The Build agent with AI (Preview) feature is an AI-assisted creation experience that generates and refines Fabric Data Agent configurations through an interactive workflow, replacing manual setup with schema- and conversation-aware recommendations for agent instructions, data source guidance, and example queries; the preview focuses on SQL and Eventhouse scenarios. For more information, see Fabric data agent concepts.
Bulk Copy API for faster ingestion (Preview) Bulk Copy API (Preview) writes data directly to Fabric Data Warehouse tables from application code without first staging files, replacing row-by-row INSERT statements to improve ingestion throughput. You can use it through SqlBulkCopy for C#, SQLServerBulkCopy for Java, the bcp utility, and tools such as SSIS, Azure Data Factory, and Informatica. For more information, see Ingest data into your warehouse by using BCP API (Preview).
Business Actions activity (Preview) Business Actions activity (Preview) lets pipelines call supported business systems and workflows. For more information, see Business Actions.
Capacity Operation Events in Real-Time Hub (Preview) Capacity Operation Events in Real-Time Hub (Preview) provide operation-level telemetry for capacity consumption, duration, throttling, workspaces, and items. For more information, see Explore Fabric capacity operation events.
Centralized data governance in the OneLake catalog (Preview) A new centralized data governance experience in the OneLake catalog is in preview. Data owners can view aggregated insights on the items they created, consider improving their governance by taking recommended actions, and access more information along with all available tools in Fabric.
Change event streaming in SQL database in Fabric (Preview) Change event streaming (CES) captures row-level DML changes (inserts, updates, and deletes) in SQL database in Fabric and publishes each change to Fabric Eventstream or Azure Event Hubs in near real time as a CloudEvent, enabling downstream analytics and event-driven architectures with no ETL. For more information, see Change event streaming overview.
Change Git branch with Contributor role (Preview) Change Git branch with Contributor role lets workspace Contributors and Members switch the connected Git branch without workspace Admin permissions. For more information, see Allow Contributors and Members to switch branches.
CI/CD Support for SQL analytics endpoint (Preview) CI/CD support for SQL analytics endpoint (Preview) lets you manage SQL analytics endpoint definitions as DacFx database projects in Git alongside other Fabric items and deploy incremental schema changes through Fabric Deployment Pipelines. For more information, see Source control (preview).
Conditional activity retries in Data Factory pipelines (Preview) Conditional activity retries (Preview) let you define exactly when a pipeline activity retries based on the error code, error message, or both with AND/OR logic, so long-running pipelines automatically recover from transient failures while fast-failing on errors that don't self-resolve. For more information, see Activity overview.
Configurable Data Retention in Fabric Data Warehouse (Preview) Configurable data retention (Preview) in Fabric Data Warehouse lets you set the data history retention window for a warehouse between 1 and 120 days with a single T-SQL command. This retention period governs time travel, point-in-time clones, restore points, and warehouse snapshots. For more information, see Data retention in Fabric Data Warehouse (Preview).
Continuous ingestion from Azure Storage to Eventhouse (Preview) You can now use continuous ingestion from Azure Storage to Eventhouse (preview) to automatically and efficiently ingest data from Azure Storage into Eventhouse. For more information, see Get data from Azure Storage.
Copilot in Fabric is available worldwide Copilot in Fabric is now available to all customers. For more information on release status, see Release status of AI and Copilot experiences in Fabric.
Copy and share warehouse SQL queries with context (Preview) You can now copy and share warehouse SQL queries with context (Preview). You can paste a query with its results or generate a link that opens the query in another tool.
Copy job and Eventstream integration (Preview) Copy job and Eventstream integration (Preview) uses Eventstream as a Copy job source or destination. For more information, see What is Copy job?.
Create AI Auto-Description for Semantic Models (Preview) Auto-Description for Semantic Models (Preview) uses Copilot to generate clear descriptions for semantic models from the model's metadata and structure, which model owners and contributors can review, edit, and regenerate directly from the semantic model details page in the OneLake catalog. For more information, see OneLake catalog overview.
Custom authorization for API for GraphQL (Preview) Authorizer User Data Functions for API for GraphQL (Preview) let you attach a User Data Function that runs before each GraphQL operation and evaluates request context such as tenant ID, role, or token claims to allow, restrict, or deny access based on your own policy logic. For more information, see What is Microsoft Fabric API for GraphQL? and Create a Fabric User data functions item.
Custom sort in Dataflow Gen2 (Preview) Custom sort in Dataflow Gen2 (Preview) defines multi-column sort precedence in one dialog. For more information, see Sort columns.
Data agent feedback in Microsoft 365 Copilot (Preview) Data agent feedback in Microsoft 365 Copilot (Preview) helps creators review consumer feedback and improve agent responses. For more information, see Data agent consumer and creator feedback.
Data agent runtime model upgrade (Preview) Data agent runtime model upgrade (Preview) uses a newer model for planning, source routing, semantic querying, and natural language to SQL. For more information, see Fabric data agent runtimes.
Data agents return up to 1,000 rows (Preview) Fabric data agents return up to 1,000 rows (Preview), expanding the results available for downstream analysis. For more information, see Data agent file support.
Data Agent support for Eventhouse functions, materialized views, and shortcuts (Preview) Data Agent now discovers and queries Eventhouse user-defined functions (UDFs), materialized views, and shortcut tables, so it calls validated KQL logic, uses pre-aggregated views for faster answers, and reaches data federated from sources outside the eventhouse. For more information, see Fabric data agent concepts.
Data agent topics (Preview) Topics in Fabric data agents (Preview) provide large, subject-specific guidance for supported data sources. For more information, see Configure Fabric data agents.
Database agent (Preview) Database agent (Preview) analyzes SQL and PostgreSQL performance, availability, workload health, and cost signals in Database Hub.
Database Hub (Preview) Database Hub (Preview) provides a unified operational experience for managing database estates across Azure, Microsoft Fabric, on-premises environments, and supported multicloud environments.
Data Factory Adaptive performance tuning (Preview) Adaptive Performance Tuning is designed to intelligently optimize data movement performance based on your configuration and runtime context. This preview feature makes tuning Data Factory simpler, safer, and more effective without requiring deep manual expertise or trial-and-error adjustments. To enable, see the adaptivePerformanceTuning property when using copy activity.
Data Factory in Fabric connection recency (Preview) Connection recency (Preview) adds last-linked-to-items and last-credentials-used properties to connections for audit and lifecycle management. For more information, see Data source management.
Data Factory MCP (Preview) Data Factory MCP (Preview) enables AI assistants to create, test, and deploy Dataflow Gen2 through natural language without manual configuration. For more information, see Data Factory MCP on GitHub.
Data Factory On-premises data gateway manual update option (Preview) The Gateway December 2025 release (version 3000.298) provides a new manual update option (in preview) via the Fabric portal or API.The November release serves as the baseline version for this feature, and you can start performing manual updates beginning in December. For more information, see Update an on-premises data gateway.
Data Factory SharePoint data source site picker (Preview) The SharePoint Site Picker (Preview) replaces manual URL entry with a browsable, searchable dropdown for selecting SharePoint sites across Dataflow Gen2, Pipelines, Copy job, and Lakehouse shortcuts. For more information, see SharePoint Folder connector.
Dataflow data visuals (Preview) Data visuals in Dataflow Gen2 (Preview) profile and visualize transformed data while you author a query. For more information, see Create data visuals in Dataflow Gen2.
Dataflow destination expressions (Preview) Dynamic destination expressions in Dataflow Gen2 (Preview) derive destination names from parameters, refresh dates, and workspace context. For more information, see Use the dynamic expression editor.
Dataflow dynamic warehouse schema (Preview) Dynamic warehouse destination schemas (Preview) evolve destination tables when Dataflow Gen2 query output changes. For more information, see Data destination schema options.
Dataflow offset transformation (Preview) Offset transformation in Dataflow Gen2 (Preview) shifts date, datetime, or time values forward or backward without custom expressions.
Dataflow refresh details in Monitoring hub (Preview) Dataflow refresh details in Monitoring hub (Preview) expose query-level refresh details without leaving the hub. For more information, see Monitor Dataflow Gen2.
Dataflow shared queries (Preview) Shared queries in Dataflow Gen2 (Preview) let colleagues publish, discover, and reuse governed Power Query logic. For more information, see My Queries and Shared Queries.
Data protection: DSPM for AI for Fabric Copilots and data agents (Preview) Data Security Posture Management (DSPM) for AI for Fabric (Preview) monitors Copilot and data agent interactions to detect sensitive information in AI prompts and responses, investigate risky AI behavior, and apply governance through Purview Audit and eDiscovery. For more information, see DSPM for AI in Fabric.
Data replication from Lakehouse with Delta Change Feed (Preview) The Fabric Lakehouse Table connector provides changed data from a Fabric Lakehouse via Delta Change Data Feed (CDF), to supported destinations. For more information, see Data Replication from Fabric Lakehouse with Delta Change Data Feed (Preview).
Data Warehouse Migration Assistant for Teradata (Preview) Get started planning your Teradata migration to Fabric Data Warehouse (preview) with the Fabric Migration Assistant for Data Warehouse.
Delegated Custom Live Pool management (Preview) Delegated Custom Live Pool management (Preview) lets workspace Members configure pools through Fabric environments when an administrator enables delegation. For more information, see Configure Custom Live Pools.
Database Hub (Preview) Database Hub (Preview) provides a unified operational experience for managing your database estate across Azure, Microsoft Fabric, on-premises environments, and supported multicloud environments.
Delegated OneLake Shortcuts (Preview) Delegated OneLake Shortcuts (Preview) let you attach an organizational account, service principal, or workspace identity to a OneLake shortcut so downstream users access the target through the delegated identity. The same model works across Microsoft Fabric tenants for parent-subsidiary and test-to-production sharing without individual cross-tenant permissions. For more information, see OneLake shortcuts and Create a OneLake shortcut.
Delta column mapping in the SQL analytics endpoint SQL analytics endpoint now supports Delta tables with column mapping enabled. For more information, see Delta column mapping. This feature is currently in preview.
DeltaFlow transformation in Fabric Real-Time Intelligence's Eventstream (preview) DeltaFlow is a capability in Fabric Eventstream that transforms raw Change Data Capture (CDC) events into a flattened, analytics-ready format. Instead of working with deeply nested Debezium JSON payloads, DeltaFlow produces tabular rows that closely mirror the structure of the source database tables, enriched with metadata columns that describe each change. Eventstream supports DeltaFlow transformations, which are currently in preview. For more information, see DeltaFlow output transformation (preview) and Business events overview.
Deployment plans (Preview) Deployment plans (Preview) define item deployment order and pre/post actions on a visual canvas. For more information, see Plan CI/CD for Fabric solutions.
Digital twin builder (Preview) Digital twin builder (preview) is an item within the Real-Time Intelligence workload. Digital twins create data-driven, real-time representations of entities. It's a data modeling item that creates digital representations of real-world environments, to optimize physical operations using data. For more information, see What is digital twin builder (preview)?
Efficient Scaledown for Spark jobs (Preview) Efficient Scaledown (Preview) decouples Spark shuffle data from executor lifetime by routing shuffle output to Azure Blob Storage, so clusters scale down faster and tolerate executor loss with no notebook or query changes. It requires the Native Execution Engine on Fabric Spark Runtime 3.5 or later. For more information, see Efficient scaledown and Native execution engine for Fabric Data Engineering.
Evaluate Power Query programmatically (Preview) Evaluate Power Query Programmatically in Microsoft Fabric (Preview) introduces a public REST API to execute Power Query M scripts programmatically, enabling automation, integration with Spark and pipelines, and access to 100+ data sources with 90-second timeout for transformations. For more information, see Execute Query API Reference.
Eventhouse workspace-monitoring logs (Preview) Eventhouse workspace-monitoring logs (Preview) expose throttling, suboptimal-size, and scale-out events. For more information, see Monitor Eventhouse.
Eventstream Business Events publisher (Preview) Eventstream now publishes Business Events as a built-in destination, so you filter, aggregate, threshold, and emit governed, discoverable business signals from the Eventstream canvas with no code.
Eventstream custom connector (Preview) Eventstream custom connector (Preview) adds proprietary or unsupported streaming sources through an extensible connector framework. For more information, see Custom stream connector.
Eventstream processing logs (Preview) Eventstream processing logs (Preview) send Azure Stream Analytics warnings, failures, and recovery events to workspace monitoring. For more information, see Eventstream workspace monitoring.
Eventstream workspace monitoring (Preview) Workspace monitoring for Eventstreams (Preview) lets you choose which eventstreams emit performance, error, and node-health data to three KQL tables in the workspace monitoring Eventhouse. For more information, see Eventstream monitoring in Workspace Monitoring and Workspace Monitoring overview.
Eventstream workspace Private Link (Preview) Eventstream workspace Private Link (Preview) secures supported streaming workloads at workspace scope.
Excel to Delta tables with shortcut transformations (Preview) Excel to Delta tables with shortcut transformations (Preview) extends the zero-code shortcut transformations ingestion model to multi-sheet .xlsx and .xls workbooks, with automatic sheet discovery, wildcard sheet matching, combined-table or table-per-sheet output, schema drift handling, continuous sync, and sanitization of sheet names into valid Delta table names. For more information, see Shortcut transformations (file).
Extended Auto partition support in Copy job for Oracle, Fabric Lakehouse, and SAP HANA (Preview) Extended Auto partition support in Copy job (Preview) automatically scales out reads and writes for Oracle, Fabric Lakehouse, and SAP HANA without predefining partition columns or custom sharding logic. For more information, see What is Copy job?
Fabric Actions activity (Preview) Fabric Actions activity (Preview) gives pipelines authenticated access to supported Fabric REST operations. For more information, see Fabric Actions.
Fabric Activator integration with warehouse SQL queries (Preview) Fabric Activator integration with warehouse SQL queries (Preview) lets you define conditions on query results and trigger follow-up actions. For more information, see Set alerts on warehouse queries.
Fabric Admin APIs Fabric Admin APIs are designed to streamline administrative tasks. The initial set of Fabric Admin APIs is tailored to simplify the discovery of workspaces, Fabric items, and user access details.
Fabric Apps backend with the Rayfin SDK (Preview) Rayfin (Preview) is an open-source SDK and CLI that lets developers and coding agents define a complete application backend in code—data models, APIs, identity, access policies, and business logic—and deploy it to Microsoft Fabric, where the resulting apps run as first-class items with built-in governance and direct access to OneLake. For more information, see Fabric Apps.
Fabric Apps functions and connectors (Preview) Fabric Apps functions and connectors (Preview) add Fabric data connectors, PostgreSQL support, server-side functions, secret storage, application storage, and built-in metrics.
Fabric-CICD tool v1.2.0 – new bulk publish mode (Preview) fabric-cicd bulk publish mode lets fabric-cicd publish multiple items in one Fabric bulk import API call and fall back to standard publishing for unsupported scenarios.
Fabric CLI in Azure DevOps (Preview) The Fabric CLI in Azure DevOps (Preview) provides the Fabric CLI (fab) as a built-in Azure DevOps pipeline task for automating workspace and item management without manual tool installation. Get started with the Fabric CLI (generally available) at https://aka.ms/FabCLI.
Fabric Connection inside Notebook (Preview) Fabric Connection inside Notebook (Preview) enables you to create and manage cloud data source connections directly within notebooks. For more information, see What is Fabric Connection in notebooks?
Fabric Data Agent integration with Azure AI Agent Service (Preview) We're excited to launch the integration of data agents in Fabric with Azure AI Agent Service from Microsoft Foundry. To get started, see Consume a Fabric Data Agent in Microsoft Copilot Studio (preview). The Fabric data agent SDK is also available in preview.
Fabric data agents evaluation with the Python SDK (Preview) You can now use the Python SDK to programmatically evaluate Fabric data agents. For more information, see Consume a Fabric Data Agent in Microsoft Copilot Studio (preview).
Fabric data engineering agent (Project Osmos) (Preview) Fabric data engineering agent (Project Osmos) runs persistent, Fabric-hosted tasks that autonomously plan, execute, and validate data engineering work against authorized lakehouse resources. Start tasks from a supported command-line client, review operational settings, and monitor progress and results in Fabric.
Fabric Data Factory Upsert table action in the Lakehouse Connector (Preview) The Upsert (Preview) table action in the Lakehouse Connector is in preview.
Fabric Data Warehouse GPU query acceleration (Preview) GPU-accelerated Fabric Data Warehouse (Early Access Preview) uses NVIDIA accelerated computing to run eligible T-SQL queries on GPUs with no query rewrites, delivering up to 7x faster performance at high concurrency across all SQL analytics endpoints and warehouses in a workspace once you enable it from workspace settings. For more information, see Query acceleration in Fabric Data Warehouse (preview).
Fabric Data Warehouse MCP server (Preview) Agentic warehouse development with MCP and Agent Skills (Preview) introduces a remote Data Warehouse MCP server that lets AI coding agents work with Fabric warehouses through authenticated, governed tools. For more information, see Data Warehouse MCP server and Skills for Microsoft Fabric.
Fabric IQ data-agent and ontology answers (Preview) Fabric IQ data-agent and ontology answers (Preview) extend business-question answering beyond reports and semantic models.
Fabric IQ (Preview) workload Fabric IQ (preview) is a new workload for unifying business semantics across data, models, and systems to power intelligent agents and decisions grounded in a live, holistic view of the business.
Fabric item relations API (Preview) Fabric item relations API (Preview) returns upstream and downstream item dependencies, typed relation edges, and related workspaces for impact analysis, catalogs, CI/CD checks, and AI agents. For more information, see Fabric REST API reference.
Fabric Maps in Real-Time Dashboards (Preview) Fabric Maps in Real-Time Dashboards (Preview) lets you add a published Fabric Map as a dashboard tile while keeping the map's layers, filters, and styling in the map item. For more information, see Add a Fabric Maps tile.
Fabric Migration Assessment for Data Factory (Preview) The new Fabric Migration Assessment for Data Factory (Preview) provides a built-in assessment and migration experience that evaluates ADF pipeline readiness and migrates supported pipelines into a Fabric workspace with connection mapping. For more information, see Migration planning for Azure Data Factory to Fabric Data Factory.
Fabric Observability Insights (Preview) Fabric Observability Insights (Preview) uses an operations agent and Fabric-native context to investigate pipeline failures. For more information, see Workspace monitoring overview.
Fabric SKU estimator (Preview) The Microsoft Fabric SKU estimator, now available in preview, is an enhanced version of the previously introduced Microsoft Fabric Capacity Calculator. For more information, see Introducing the Microsoft Fabric SKU estimator (preview) and Mastering SKU Estimations with the Microsoft Fabric SKU Estimator.
File-level Git commit (Preview) File-level Git commit (Preview) commits selected files from a supported Fabric item while unfinished files remain in the workspace. For more information, see Compare and commit Fabric items.
Folder REST API (Preview) You can now create and manage workspace folders in automation scenarios and integrate with other systems and tools. The Folder Rest API is now in preview. To get started, see Fabric REST API Folders.
Git Integration Workspace Relation API (Preview) The Workspace Relations API lets automation create, query, update, and delete relationships between parent and branched workspaces. For more information, see Development process using the Branch out experience.
Google Lakehouse Runtime Catalog mirroring (Preview) Google Lakehouse Runtime Catalog mirroring (Preview) brings Google-managed Iceberg catalog metadata into OneLake without copying data, so you can use the tables across Fabric workloads. For more information, see Google Lakehouse Runtime Catalog mirroring.
Improved NL2SQL Engine for Fabric Data Agent (Preview) Improved NL2SQL Engine (Preview) for Fabric Data Agent improves accuracy, transparency, and resilience when translating natural language to SQL across lakehouses, warehouses, and mirrored databases. It uses example queries to follow patterns, asks clarifying questions when intent is ambiguous, and surfaces structured diagnostics, available when you enable the Preview Fabric Data Agent Runtime. For more information, see Fabric data agent concepts.
Investigator insights in Operations Agent (Preview) Investigator insights in Operations Agent analyze related data after an operations agent alert and surface scope, observations, patterns, and recommended next steps in Teams. For more information, see Operations Agent actions.
Invoke SSIS Package activity in Data Factory pipelines (Preview) The Invoke SSIS Package activity (preview) enables package execution from pipeline orchestration in Fabric. For more information, read Blog: Invoke SSIS Package Activity in Microsoft Fabric (preview).
Item History in Fabric Capacity Metrics App (Preview) The Item History page in the Fabric Capacity Metrics App (Preview) provides a 30-day view of compute capacity consumption with interactive visuals and slicers for workspace and item-level analysis. For more information, see Understand the metrics app item history page (preview).
Item soft-delete and recovery events (Preview) Item soft-delete and recovery events (Preview) expose lifecycle events in Real-Time Hub for automation and audit scenarios. For more information, see Fabric workspace item events.
JobInsight Diagnostics Library (Preview) JobInsight is a diagnostics library to analyze completed Spark applications via APIs for queries, jobs, stages, tasks, executors, and event logs. For more information, see Job Insight Library (Preview).
Lakehouse auto-binding in Git (Preview) Lakehouse auto-binding in Git (Preview) resolves lakehouse references automatically as notebooks move across Git-connected workspaces, making notebooks portable across development, test, and production environments. For more information, see Notebook source control and deployment.
Lakehouse support for git integration and deployment pipelines (Preview) The Lakehouse now integrates with the lifecycle management capabilities in Microsoft Fabric, providing a standardized collaboration between all development team members throughout the product's life. Lakehouse git integration and deployment pipelines facilitates an effective product versioning and release process by continuously delivering features and bug fixes into multiple environments.
Large Eventstream and Eventhouse messages (Preview) Large message support (Preview) ingests richer event payloads through Eventstream and Eventhouse. For more information, see Eventstream overview.
Load Fabric OneLake Data in Excel Easily load Fabric OneLake data into Excel with integrated OneLake catalog and modern Get Data experience (preview). For more information and steps to get started, see OneLake catalog and Get Data are integrated into Excel for Windows.
Manage failure notifications from the monitoring hub in Fabric (Preview) The Schedule failures page in the Fabric monitoring hub (Preview) lists every scheduled item with failure notifications configured and lets you add, edit, or remove recipients for any item from one place, without opening each item individually. For more information, see Schedule failures (preview) and Receive notifications for failed scheduled jobs.
Mapping Data Flow Transforms in Dataflow Gen2 (Preview) Mapping data flow transforms in Dataflow Gen2 (Preview) introduces Spark-based transformation capabilities directly in the Dataflow Gen2 creation experience, enabling large-scale data transformations with Spark compute and migration of mapping data flows to Fabric. For pricing details, see Dataflow Gen2 pricing.
Maps external feature services (Preview) External feature services in Fabric Maps (Preview) connect hosted GIS and standards-based geospatial services. For more information, see Create layers from external feature services.
Materialized Lake View refresh features (Preview) The Refresh Materialized Lake View activity in Fabric Data Factory pipelines in Fabric Data Factory pipelines lets you refresh materialized lake views on demand or on a schedule within your pipeline orchestration. Event-driven refresh for materialized lake views triggers dependent materialized lake view refreshes when OneLake file events or pipeline and notebook job events occur. For more information, see Schedule a materialized lake view refresh.
Maven support in Fabric Environment (Preview) Maven support in Fabric Environment (Preview) lets Scala and Java developers add libraries from Maven repositories to a Fabric Environment by uploading and managing pom.xml files directly, so transitive dependencies are resolved without manually downloading and uploading JARs. For more information, see Manage libraries in Fabric environments.
MAX_ROLLOVER_FILES in extended event sessions using Azure Storage blobs (preview) For new extended event sessions in SQL database in Fabric, you can specify the MAX_ROLLOVER_FILES option to retain only the specified number of the most recent Azure Storage blobs. For more information, see event_file target.
Max vCores limit option in SQL databases (Preview) The Max vCores limit option (Preview) lets you cap SQL database compute at 4 or 32 vCores to control costs and protect shared capacities from noisy-neighbor scenarios. For more information, see Control compute usage.
MCP remote servers for Activator and Eventhouse in Real-Time Intelligence (preview) Fabric Real-Time Intelligence provides hosted Model Context Protocol (MCP) remote servers for Activator and Eventhouse, enabling AI assistants and agents to interact with RTI components through natural language. By using the Activator MCP server, AI assistants can create monitoring rules, manage alerts, and trigger actions in Fabric Activator. The Eventhouse remote MCP server allows AI agents to query, reason over, and act on real-time data stored in an RTI eventhouse by configuring a URL that points to the Eventhouse. For more information, see Get started with the remote MCP server for eventhouses (preview) and Get started with the remote MCP server for Activator (preview). This feature is currently in preview.
Metropolis VoiceLake (Preview) Metropolis VoiceLake (Preview) brings supported communications data into a lakehouse with a semantic model and Power BI reports.
Microsoft ADO.NET Driver for Fabric Data Engineering (Preview) The Microsoft ADO.NET Driver for Fabric Data Engineering (Preview) lets .NET applications connect to Spark SQL in Fabric using standard ADO.NET patterns with Microsoft Entra ID authentication, connection pooling, and async prefetch via Livy APIs. For more information, see Microsoft ADO.NET driver for Microsoft Fabric Data Engineering (preview).
Microsoft Fabric Embed (Preview) Microsoft Fabric Embed (Preview) lets you embed Real-Time Dashboards in JavaScript or TypeScript apps with delegated Microsoft Entra access. For more information, see Power BI embedded analytics and What is Microsoft Fabric Embed in Microsoft Fabric (Preview)?.
Microsoft Fabric policies (Preview) Microsoft Fabric policies let administrators centrally define and enforce condition-based governance rules at tenant or capacity scope. Policies support centralized management in the OneLake catalog, audit logging, REST APIs, Git integration, and CI/CD automation.
Microsoft ODBC Driver (Preview) The Microsoft ODBC Driver for Fabric Data Engineering (Preview) enables .NET, Python, and other ODBC-compatible applications and BI tools to connect to Spark SQL in Fabric through Livy APIs with Microsoft Entra ID authentication, session reuse, and async prefetch. For more information, see Microsoft ODBC driver for Microsoft Fabric Data Engineering (Preview).
Mirrored Dremio catalog in OneLake (Preview) Fabric can mirror Dremio Iceberg REST Catalog metadata to enable zero-copy access to Dremio-managed Iceberg tables in OneLake via shortcuts. Tables appear in OneLake within seconds after selection, with optional auto-include of newly added tables and SQL analytics endpoint querying across Fabric workloads. For more information, see Bring your Dremio data into OneLake (Preview) and Mirrored Dremio catalog.
Mirroring for Azure Database for MySQL (Preview) Mirroring in Fabric provides an easy experience to avoid complex ETL (Extract, Transform, Load) processes and seamlessly integrate your existing Azure Database for MySQL data with the rest of your data in Fabric. For more information, see Mirroring Azure Database for MySQL.
ML endpoint runtime and tensor support (Preview) ML endpoint runtime and tensor support (Preview) add Spark 4.1 and tensor-based model support for real-time serving scenarios.
Modern evaluation engine for VNet data gateways (Preview) Modern evaluation engine for VNet data gateways (Preview) provides an opt-in execution path for supported mashup evaluations so you can test semantic model refresh, Dataflow Gen2 refresh, and DirectQuery performance without changing your private network architecture. For more information, see Modern evaluation engine for VNet data gateways.
Modern Get Data Azure browsing (Preview) Azure resource browsing in Modern Get Data (Preview) selects supported Azure resources without copying endpoint values from the Azure portal. For more information, see Get data in Dataflow Gen2.
Monitor for Fabric Data Warehouse (Preview) Monitor for Fabric Data Warehouse (Preview) (previously Query Activity) is unified monitoring for Fabric Data Warehouse that shows active and historical queries side by side, supports cross-execution performance analysis powered by Query Insights, and lets you cancel long-running queries with one click. For more information, see Monitor T-SQL queries in Fabric Data Warehouse (Preview).
MSSQL extension for VS Code Fabric integration (Preview) MSSQL extension for VS Code Fabric integration (Preview) adds support for connecting, running queries, and managing objects in SQL databases in Fabric directly from Visual Studio Code. Download the extension at MSSQL extension at marketplace.visualstudio.com.
Multiple-Schema Inferencing in Eventstream (Preview) Multiple schema inferencing (preview) lets you work with multiple data sources that emit varying schemas by inferring and managing multiple schemas simultaneously. For more information, see Enhancing Data Transformation Flexibility with Multiple-Schema Inferencing in Eventstream (Preview).
Native Execution Engine JSON acceleration (Preview) Native Execution Engine JSON acceleration (Preview) reads and parses JSON through the vectorized native path so supported Spark transformations can stay columnar without code changes. For more information, see Native execution engine for Fabric Data Engineering and Apache Spark runtime in Fabric.
Native Planning Engine (Preview) Native Planning Engine (Preview) scales planning models to tens of millions of cells and recalculates dependent values as inputs change. For more information, see Planning in Fabric.
Notebook Copilot inline code completion (Preview) Now in preview, Copilot Inline Code Completion (Preview) is an AI feature that assists data scientists and engineers in writing Python code more quickly and easily. For more information, see Notebook Copilot inline code completion (Preview).
Notebook debug within vscode.dev (Preview) You can now place breakpoints and debug your Notebook code with the Synapse VS Code - Remote extension in vscode.dev. This update first starts with the Fabric Runtime 1.3 (EOSA).
Notebook toolkit for agentic development (Preview) Fabric Notebook Toolkit (Preview) gives coding agents notebook-aware tools for authoring and validation.
Observability for Fabric Data Agent in Microsoft Foundry (Preview) Observability for Fabric Data Agent in Microsoft Foundry (Preview) surfaces telemetry for every call to a Fabric Data Agent used as a tool inside a Foundry agent, including latency, status, and error details, so agent builders can debug and monitor production agents. For more information, see Fabric data agent concepts.
OneLake Catalog search API, MCP, and CLI tools (Preview) OneLake Catalog search API and MCP tool (Preview) brings cross-workspace discovery to code with a single search request to locate matching items across your accessible estate based on catalog metadata and user permissions, with the capability also included as a built-in tool in the Fabric Core MCP server for AI agents. The Fabric CLI fab find command extends the search to the terminal so you can locate items across workspaces, filter or exclude by item type, and pipe results through JMESPath. For more information, see OneLake catalog overview.
OneLake Catalog table discovery (Preview) OneLake Catalog table discovery (Preview) lets you find tables from semantic models, lakehouses, and mirrored databases by table name, description, or exact column name in global search and the Search API. For more information, see OneLake Catalog Search API and Fabric Core remote MCP server.
OneLake table read API (Preview) Use the table read API to read rows from a Delta Lake or Apache Iceberg table in OneLake. For more information, see Overview of OneLake table APIs.
On-premises data gateway manual update (Preview) Manual update for the on-premises data gateway is now in preview, allowing administrators to trigger gateway updates through the Fabric portal, API, or PowerShell scripts on their own schedule. For more information, see Update an on-premises data gateway.
On-premises data gateway soft delete (Preview) On-premises data gateway soft delete (Preview) retains supported standard mode gateways for 30 days by default so tenant and gateway admins can recover accidental deletions. For more information, see Gateway soft delete.
Ontology as data-agent context (Preview) Ontology as data-agent context (Preview) grounds query generation in enterprise entities, relationships, definitions, and source bindings. For more information, see Ontology as context.
Ontology as Real-Time Dashboard source (Preview) Ontology as a Real-Time Dashboard source (Preview) brings business entities and relationships into live visualizations.
Ontology (preview) item Ontology (preview) in Fabric IQ lets you define entity types, relationships, properties, and other constraints to organize data according to your business vocabulary. For more information, see What is ontology (preview)?
OpenAI plugins for Eventhouse (Preview) You can now use two powerful AI plugins for Eventhouse: AI Embed Text Plugin and AI Chat Completion Prompt Plugin. Connect Eventhouse data to OpenAI-powered applications for advanced analytics and AI scenarios. For more information, see ai_embed_text (preview) and ai_chat_completion (preview).
Operations Agent for Pipelines (Preview) Operations Agent for Pipelines is an AI-powered monitoring agent in Data Factory that continuously watches your pipeline runs and proactively notifies you through Microsoft Teams when failures or anomalies occur. You can configure the agent to monitor specific scenarios such as all pipeline failures or runs exceeding expected durations, without writing custom alerting logic. For more information, see Operations Agent for Pipelines (preview).
Operations Agent performance view (Preview) Operations Agent performance view (Preview) shows monitoring, reasoning, and model-usage telemetry for an operations agent.
Optimal refresh for materialized lake view updates and deletes (Preview) Optimal refresh for updates and deletes lets materialized lake views process changed and deleted source rows incrementally. Refresh hints identify unique rows so Fabric can avoid a full refresh when supported source data changes.
Oracle Database CDC source connector for Fabric Eventstream (Preview) The Oracle Database CDC source connector for Eventstream (Preview) streams changes from an Oracle database into Fabric as a structured change event with before and after values, using Oracle LogMiner for on-premises or cloud databases running Oracle 12c and later. For more information, see Add Oracle Database CDC source to an eventstream (preview).
Pagination support for Eventstream HTTP connector (Preview) You can add an HTTP source to an eventstream (preview) to ingest paginated REST API responses through page-based or cursor-based pagination, eliminating custom orchestration when bringing activity logs, audit records, or operational events into Eventstream.
Partitioned compute for Dataflow Gen2 (Preview) Partitioned compute is a capability of the Dataflow Gen2 engine that allows parts of your dataflow logic to run in parallel, reducing the time to complete its evaluations. For more information, see partitioned compute in Dataflow Gen2.
Pass Parameter Values to Fabric Items (Preview) Activator enables you to automatically activate Fabric items like pipeline and notebook whenever certain data conditions are met. You can not only activate and execute Fabric items but also pass values to the parameters defined in your Fabric items.
Pipeline-level dependencies (Preview) Pipeline-level dependencies (Preview) run downstream pipelines only when configured upstream conditions are met. For more information, see Pipeline run conditions.
Prebuilt Foundry Tools in Fabric preview The preview of prebuilt Foundry Tools in Fabric is an integration with Foundry Tools, formerly known as Azure AI services. Foundry Tools (preview) allow for easy enhancement of data with prebuilt AI models without any prerequisites. Currently, prebuilt Foundry Tools are in preview and include support for the Microsoft Azure OpenAI Service, Azure Language in Foundry Tools, and Azure Translator in Foundry Tools.
Public API capabilities for Dataflow Gen2 dataflows in Fabric Data Factory (Preview) This preview of the Dataflows Gen 2 public APIs enables users to create, update, and monitor their data workflows programmatically. The APIs support a wide range of operations including dataflows CRUD (Create, Read, Update, and Delete), scheduling, and monitoring, making it easier for users to manage their data integration processes.
Public parameter values to refresh a Dataflow Gen2 (Preview) Learn more about the new public parameters capability for Dataflow Gen2 with CI/CD support, and the support for this new mode within the Dataflow refresh activity in Data Pipelines.
Real-Time Dashboard AI custom visual builder (Preview) AI custom visual builder (Preview) generates purpose-built HTML visuals for Real-Time Dashboards from natural-language descriptions. For more information, see AI custom visual in Real-Time Dashboards.
Real-Time Dashboards Time Series Visualization (Preview) Real-Time Dashboard-specific visuals in Real-Time Dashboards adds dedicated capabilities for navigating, comparing, and customizing time-based data, including a legend search bar, pin-and-overlay for series comparison, multiple aligned-time panels, flexible Y-axis scaling, color assignment, and zoom with a time slider.
Recent data in Dataflow Gen2 (Preview) Recent data in Dataflow Gen2 (Preview) provides quick access to your most frequently used data items in the Power Query ribbon and Modern Get Data, letting you jump directly to recently used tables, files, folders, and databases without navigation. For more information, see Get data experience and What is Dataflow Gen2?
Reference data enrichment in Eventstream (Preview) Reference data enrichment in Eventstream joins streaming events with Delta reference tables from a lakehouse by using no-code or SQL-based enrichment. For more information, see Reference data join in Eventstream using Lakehouse.
Refreshed Monitor hub (Preview) Refreshed Monitor hub (Preview) adds at-scale health views across workloads, jobs, apps, agents, alerts, and capacities. For more information, see Monitor hub.
Refresh SQL analytics endpoint activity (Preview) The Refresh SQL analytics endpoint activity (Preview) refreshes a lakehouse SQL analytics endpoint on demand or on a schedule within a Fabric pipeline.
Resource Profiles in Fabric Data Engineering (Preview) Resource Profiles in Fabric Data Engineering (Preview) are preconfigured Spark compute profiles—write-heavy, read-heavy for Spark, and read-heavy for Power BI—that you select per environment so Fabric applies workload-aware settings automatically without manual tuning of dozens of Spark properties. For more information, see Configure Resource Profile Configurations.
REST APIs for connections and gateways (Preview) REST APIs for connections and gateways are now in preview. These new APIs allow developers to programmatically manage and interact with connections and gateways within Fabric.
Scalar user-defined functions (UDFs) Scalar user-defined functions (UDFs) are now supported as a preview feature. For more information, see CREATE FUNCTION for Fabric Data Warehouse.
Scheduled User Data Functions (Preview) Scheduled User Data Functions (Preview) run parameterized business logic on a recurring cadence through the Fabric job scheduler, with execution history available in the monitoring hub. For more information, see User Data Functions overview and Schedule a User Data Function.
Send email from Dataflow Gen2 (Preview) Send email from Dataflow Gen2 (Preview) sends data-driven email directly from a dataflow without a separate orchestration step.
Service principal support for Fabric data agents (Preview) Fabric data agents now support service principal (SPN) authentication, so custom apps and Microsoft Foundry agents call the data agent API with an application identity instead of a delegated user token. The data agent inherits workspace permissions.
Snowflake security-role mirroring (Preview) Snowflake security-role mirroring (Preview) brings Snowflake role definitions into Fabric. For more information, see Extended capabilities in Mirroring.
SQL analytics endpoint metadata sync option (preview) You can now opt in to the new SQL analytics endpoint metadata sync option, a preview feature. The new option, which is currently opt-in only and currently applies only to new SQL analytics endpoints, is faster, provides a new manual metadata sync option, and provides time travel querying.
SQL database data virtualization (Preview) Data virtualization (preview) in SQL database enables querying external data stored in OneLake using T-SQL. With data virtualization syntax, you can execute Transact-SQL (T-SQL) queries on files that store data in common data formats in OneLake. You can combine this data with locally stored relational data by using joins.
SQL database skills (Preview) SQL database skills (Preview) give AI coding agents workload-specific guidance for authoring schemas, querying data, and operating SQL databases as part of broader Fabric solutions. For more information, see Work with SQL database skills in Microsoft Fabric.
SQL database support for private links (Preview) You can use tenant and workspace-level private links to provide secure access for data traffic in Microsoft Fabric, including SQL database. For SQL database, workspace-level private links are in preview. For more information, see Set up and use private links.
Stream Mirrored Database change feeds into Eventstreams (Preview) Mirrored Database change feed connector for Eventstreams (Preview) streams Delta Change Data Feed (CDF) row-level inserts, updates, and deletes from a mirrored database directly into a Fabric Eventstream for low-latency, event-driven processing, with full source schema and change metadata fidelity and support for Azure SQL, Cosmos DB, Oracle, PostgreSQL, Snowflake, and Open Mirroring partners. For more information, see Extended capabilities in mirroring for Fabric.
Time Travel using SQL analytics endpoint (Preview) Time Travel using SQL analytics endpoint (Preview) extends the Fabric Data Warehouse time travel experience to the SQL analytics endpoint so you can query Lakehouse data exactly as it looked at any prior point in time by adding OPTION (FOR TIMESTAMP AS OF '...') once at the T-SQL statement level.
Update policies on accelerated shortcuts (Preview) Update policies on accelerated shortcuts (Preview) apply ingestion-time transformations to query-accelerated OneLake shortcuts. For more information, see Update policies on accelerated shortcuts.
Upgrade Dataflow Gen1 to Dataflow Gen2 with the Upgrade Wizard (Preview) The Dataflows Upgrade Wizard upgrades one or more Dataflow Gen1 items in place while retaining their IDs, names, schedules, and connections. For more information, see Upgrade Dataflow Gen1 to Dataflow Gen2 using the Upgrade Wizard.
Upsert to delta table with Lakehouse connector (Preview) We've added upsert support to the Lakehouse connector, allowing direct writes to Delta tables, in both Copy job and Copy activity within Pipeline. For more information, see Configure Lakehouse in a copy activity.
Warehouse AI functions (preview) Fabric Data Warehouse now brings AI directly into T-SQL, enabling you to categorize and classify text, analyze sentiment, extract structured information, translate text across languages, and even correct grammar - all using built-in AI functions. To get started, see Use AI functions in Fabric Data Warehouse (preview).
Warehouse analytical functions (Preview) Warehouse analytical functions (Preview) add built-in statistical calculations to Fabric Data Warehouse and the SQL analytics endpoint. For more information, see Aggregate functions.
Warehouse CI/CD 2.0 (Preview) Warehouse item definition version 2.0 uses DacFx-based incremental extraction for Git integration and more accurate comparisons and deployment plans in deployment pipelines. For more information, see Source control with Warehouse and Upgrade Fabric Data Warehouse system files.
Warehouse custom SQL pools (Preview) Using custom SQL pools to balance performance and cost explains how workload routing and pool allocation affect performance, concurrency, and capacity consumption. For more information, see Configure custom SQL pools.
Warehouse data clustering (Preview) Data clustering is a technique used to organize and store data based on similarity. Data clustering improves query performance and reduces compute and storage access costs for queries by grouping similar records together. For more information and to get started, see Data clustering in Fabric Data Warehouse and Use data clustering in Fabric Data Warehouse (Preview).
Warehouse pre/post deployment support (Preview) Warehouse pre/post deployment support (Preview) runs validation and setup scripts around automated warehouse deployments.
Warehouse source control (Preview) Using Source control with Warehouse (preview), you can manage development and deployment of versioned warehouse objects. You can use SQL Database Projects extension available inside of Visual Studio Code. For more information on warehouse source control, see CI/CD with Warehouses in Microsoft Fabric.
Warehouse T-SQL enhancements (Preview) Warehouse T-SQL enhancements (Preview) add constructs such as QUALIFY and expanded grouping syntax. For more information, see QUALIFY.
Warehouse User Data Function integration (Preview) Warehouse User Data Function integration (Preview) invokes custom Python logic from T-SQL external functions. For more information, see Fabric User Data Functions in Warehouse.
Workspace-level surge protection (Preview) Workspace-level surge protection lets capacity admins cap a workspace's compute consumption, automatically or manually block excessive consumers, exempt mission-critical workspaces, and restore blocked workspaces. For more information, see Manage surge protection for Fabric capacities.
Workspace monitoring (Preview) Workspace monitoring uses a monitoring item to collect logs and metrics from supported Fabric items. Centralize telemetry from multiple workspaces, query it with KQL, build dashboards and reports, create alerts, and use Operations Agent to investigate issues.
Workspace outbound access protection for Azure and Fabric events (Preview) Workspace outbound access protection (OAP) for Azure and Fabric events (Preview) extends outbound network controls to cross-workspace event consumption. For more information, see Tenant private links for Azure and Fabric events.
Workspace outbound access protection for Fabric Maps (Preview) Workspace outbound access protection for Fabric Maps (Preview) allows only approved map data sources and external geospatial services through workspace data connection rules. For more information, see Workspace outbound access protection for Fabric Maps.
Workspace outbound access protection for Operations Agent (Preview) Workspace outbound access protection for Operations Agent applies workspace policies to outbound actions such as Teams notifications, workflow triggers, and cross-workspace operations. For more information, see Workspace outbound access protection for Operations Agent.
Workspace outbound access protection for Real-Time Intelligence (Preview) Workspace outbound access protection (OAP) for Real-Time Intelligence (Preview) extends workspace-level outbound network controls to Eventstream, Eventhouse, KQL QuerySet, Activator, Real-Time Dashboards, and Azure and Fabric events, so admins block unwanted outbound connections by default and allow only approved destinations through data connection rules. For more information, see Workspace outbound access protection.
Zero-copy access to OneLake data in Azure Databricks (Preview) OneLake catalog federation in Azure Databricks (Preview) enables Unity Catalog to query Fabric data stored in OneLake without copying, keeping OneLake as the source of truth while allowing Databricks compute to analyze tables through synchronized metadata and zero-copy data access. For more information, see Enable OneLake catalog federation.

Generally available features

The following table lists the features of Microsoft Fabric that recently transitioned to general availability (GA).

Month Feature Learn more
September 2026 Fabric IQ in Microsoft 365 Copilot (Generally Available) Fabric IQ in Microsoft 365 Copilot Chat and Copilot Cowork answers business questions with governed Power BI reports and semantic models. For more information, see Fabric IQ in Microsoft 365 Copilot.
September 2026 SQL CDC custom capture names (Generally Available) SQL CDC custom capture names let Copy job use existing CDC tables that don't follow the default naming convention. For more information, see Change data capture in Copy job.
September 2026 Copy job workspace monitoring (Generally Available) Copy job workspace monitoring centralizes execution status, throughput, and health trends. For more information, see Monitor a Copy job.
September 2026 Custom Copy job staging (Generally Available) Custom Copy job staging selects the staging database and schema for CDC and SCD Type 2 scenarios. For more information, see Change data capture in Copy job.
September 2026 Snowflake read-only CDC (Generally Available) Snowflake read-only CDC lets Copy job ingest changes from read-only Snowflake databases. For more information, see Snowflake CDC in Copy job.
September 2026 Oracle CDC initial-load improvements (Generally Available) Oracle CDC initial-load improvements improve onboarding performance and scale across multiple source tables.
September 2026 PostgreSQL connector for Copy job (Generally Available) Azure Database for PostgreSQL connector supports full and watermark-based incremental copy. For more information, see Copy job connectors.
September 2026 Multi-folder Copy job (Generally Available) Multi-folder Copy job loads multiple source folders into separate destination tables. For more information, see Copy job connectors.
September 2026 SharePoint folders in Copy job (Generally Available) SharePoint folders in Copy job use SharePoint folders as sources or destinations. For more information, see Copy job connectors.
September 2026 Optimized Dataflow copy to lakehouse (Generally Available) Optimized Dataflow copy to lakehouse reduces serialization overhead when Dataflow Gen2 loads staged data into lakehouse tables. For more information, see Staged data options.
September 2026 My Queries in Dataflow Gen2 (Generally Available) My Queries saves personal Power Query logic for reuse across dataflows. For more information, see My Queries and Shared Queries.
September 2026 New Dataflow destinations (Generally Available) New Dataflow Gen2 destinations write prepared data to Excel workbooks and Snowflake. For more information, see Excel workbook destinations.
September 2026 Dataflow diagnostics download (Generally Available) Dataflow diagnostics download provides structured run logs for troubleshooting and support. For more information, see Monitor Dataflow Gen2.
September 2026 Connection authentication allowlist (Generally Available) Connection authentication allowlist centrally restricts allowed authentication methods. For more information, see Authentication allowlist.
September 2026 Connection tenant allowlist (Generally Available) Connection tenant allowlist limits cloud connections to trusted Microsoft Entra tenants. For more information, see Tenant allowlist.
September 2026 Cloud connection Admin APIs (Generally Available) Cloud connection Admin APIs inventory tenant connections, identify orphaned connections, and support governance automation. For more information, see Connection Admin APIs.
September 2026 Atlan connector for Fabric (Generally Available) Atlan connector for Fabric adds business context and deeper lineage across Fabric items.
September 2026 ArcGIS Maps for Fabric (Generally Available) ArcGIS Maps for Microsoft Fabric adds spatial analysis, stream layers, and OneLake mapping to Fabric.
September 2026 Telmai Data Reliability (Generally Available) Telmai Data Reliability for Fabric monitors OneLake tables for volume, schema, freshness, and completeness changes.
September 2026 Celonis Process Intelligence (Generally Available) Celonis Process Intelligence brings process flows and root-cause analysis into Fabric.
September 2026 Spectral Core Fabric Migration (Generally Available) Spectral Core Fabric Migration guides database movement, SQL translation, schema conversion, and validation.
September 2026 Financial Fabric DataHub (Generally Available) Financial Fabric DataHub brings capital-markets datasets and conversational exploration into Fabric.
September 2026 IntuigenceAI for Fabric (Generally Available) IntuigenceAI for Fabric turns industrial documents and logs into a queryable knowledge graph in OneLake.
September 2026 Capacity Metrics app enhancements (Generally Available) Capacity Metrics app enhancements add health-state duration, capacity-units-per-second views, heatmaps, contributor drilldowns, and chargeback integration. For more information, see Capacity Metrics app.
September 2026 OneLake catalog Govern experience (Generally Available) OneLake catalog Govern experience centralizes governance for tenants, domains, capacities, workspaces, tags, and policies. For more information, see Govern your data estate.
September 2026 Materialized lake views (Generally Available) Materialized lake views provide reusable, managed lakehouse transformations for analytics.
September 2026 Capacity alert templates (Generally Available) Capacity alert templates create common capacity-health alerts from ready-made patterns.
September 2026 Event Schema Registry (Generally Available) Event Schema Registry defines and validates contracts for reliable event producers and consumers. For more information, see Schema Registry overview.
September 2026 New Eventstream connectors (Generally Available) New Eventstream connectors add Cribl, Solace PubSub+, SAP Datasphere, MongoDB CDC, and Mirrored Database Change Feed sources. For more information, see Add a Cribl source and Add a Solace PubSub+ source.
September 2026 Activator query visibility (Generally Available) KQL and SQL query visibility in Activator shows the queries behind supported alerts and actions. For more information, see Create Activator alerts from a KQL Queryset.
September 2026 Activator Fabric-item actions (Generally Available) Activator Fabric-item actions start Copy jobs, Dataflows, and User Data Functions from Activator rules. For more information, see Trigger Fabric items.
September 2026 Activator stateful conditions (Generally Available) Activator stateful conditions evaluate changes over time instead of treating each event independently. For more information, see Activator rules overview.
September 2026 Granular shortcut-database sharing (Generally Available) Granular shortcut-database sharing shares selected Eventhouse data without exposing an entire database. For more information, see OneLake shortcuts.
September 2026 Eventhouse entity diagram (Generally Available) Eventhouse entity diagrams visualize KQL database entities and relationships. For more information, see View an entity diagram.
September 2026 Eventhouse outbound access protection (Generally Available) Eventhouse outbound access protection blocks outbound traffic by default and permits approved destinations through data connection rules. For more information, see Workspace outbound access protection.
September 2026 Eventhouse workspace Private Link (Generally Available) Eventhouse workspace Private Link secures supported ingestion, querying, monitoring, and reuse within private-link-enabled workspaces. For more information, see Tenant and workspace Private Links.
September 2026 Eventhouse customer-managed keys (Generally Available) Eventhouse customer-managed keys encrypt dedicated physical clusters with organization-controlled Azure Key Vault keys. For more information, see Eventhouse encryption.
September 2026 Fabric IQ MCP Server (Generally Available) Fabric IQ MCP Server brings governed Fabric business context to compatible AI clients. For more information, see Fabric IQ MCP Server.
September 2026 Fabric data agent visualizations (Generally Available) Fabric data agent visualizations return interactive charts with text and tables. For more information, see Get visual responses from a data agent.
September 2026 Data-agent MCP tasks (Generally Available) Data-agent MCP tasks support long-running Model Context Protocol operations through a published data-agent endpoint. For more information, see Data agent as an MCP server.
September 2026 Lakehouse maintenance activity (Generally Available) Lakehouse maintenance activity runs VACUUM and OPTIMIZE operations as orchestrated pipeline steps. For more information, see Lakehouse maintenance activity.
September 2026 Pipeline retry backoff (Generally Available) Pipeline retry backoff progressively increases the interval between activity retry attempts. For more information, see Activity retries.
September 2026 Extended mirroring capabilities (Generally Available) Extended mirroring capabilities add change feeds and source-view mirroring for advanced replication scenarios. For more information, see Extended capabilities in Mirroring.
September 2026 Copy job CDC (Generally Available) Change data capture in Copy job incrementally processes inserted, updated, and deleted source records. For more information, see Change data capture in Copy job.
September 2026 Copy job SCD Type 2 (Generally Available) SCD Type 2 in Copy job preserves historical record versions for supported CDC scenarios. For more information, see Change data capture in Copy job.
September 2026 Copy job audit columns (Generally Available) Copy job audit columns add operational metadata to destination rows, including Snowflake destinations. For more information, see Audit columns in Copy job.
September 2026 DLP restrict access (Generally Available) DLP restrict access (Generally Available) automatically limits access when Microsoft Purview policies detect sensitive information. For more information, see Restrict access action for DLP policies.
September 2026 Fabric Core MCP Server (Generally Available) Fabric Core MCP Server gives compatible AI agents authenticated, audited access to Fabric workspaces, items, permissions, search, and capacity information. For more information, see Get started with Fabric Core MCP Server.
September 2026 Branch workspaces and selective branching (Generally Available) Branch workspaces and selective branching create governed, isolated workspaces with only the items developers need. For more information, see Development process using branch workspaces.
September 2026 Compare and commit Git changes (Generally Available) Compare and commit Git changes provides scoped diffs, conflict resolution, and item commits. For more information, see Compare and commit Fabric items.
September 2026 Bulk item definition APIs (Generally Available) Bulk item definition APIs export or import multiple Fabric item definitions in one operation. For more information, see Items API reference.
September 2026 OneLake outbound access protection (Generally Available) OneLake outbound access protection blocks external shortcuts and copy operations by default and allows approved connections. For more information, see Manage outbound access.
September 2026 OneLake security Members and Data UX (Generally Available) OneLake security Members and Data experiences show role assignments and the data each role can access. For more information, see Create and manage OneLake security roles.
September 2026 Custom Live Pools (Generally Available) Custom Live Pools keep scheduled Spark clusters ready with custom libraries for predictable notebook startup. For more information, see Custom Live Pools overview.
September 2026 Spark connector for SQL databases (Generally Available) Spark connector for SQL databases supports production reads and bulk writes across supported SQL platforms. For more information, see Spark connector for SQL databases.
September 2026 Fabric Livy API high concurrency (Generally Available) Fabric Livy API high concurrency runs multiple isolated Spark workloads in shared sessions. For more information, see High concurrency in the Fabric Livy API.
September 2026 ADO.NET driver high concurrency (Generally Available) ADO.NET driver high concurrency runs concurrent Spark SQL queries through shared sessions. For more information, see ADO.NET driver high concurrency.
September 2026 JDBC driver high concurrency (Generally Available) JDBC driver high concurrency runs concurrent Spark SQL workloads from Java applications. For more information, see JDBC driver for Fabric Data Engineering.
September 2026 ODBC driver high concurrency (Generally Available) ODBC driver high concurrency runs concurrent Spark SQL workloads on Windows and Linux. For more information, see ODBC driver for Fabric Data Engineering.
September 2026 Lakehouse Explorer resizable columns (Generally Available) Resizable Lakehouse Explorer columns adjust table-preview widths for long names and values. For more information, see Navigate the Lakehouse explorer.
September 2026 SQL pool statement classifier (Generally Available) SQL pool statement classification automatically separates data-loading and query statements into workload pools. For more information, see Custom SQL pools.
September 2026 Distributed bitmap filters (Generally Available) Distributed bitmap filters reduce data movement for supported fact-to-dimension joins in Fabric Data Warehouse.
September 2026 Business Events (Generally Available) Business Events publish governed business signals for independent operational and analytical consumers, and a default eventhouse automatically retains every published event. For more information, see Business events overview.
September 2026 Richer Real-Time Dashboard visuals (Generally Available) Richer Real-Time Dashboard visuals add responsive KPI modes, reference lines, titles, text, and dividers. For more information, see Customize Real-Time Dashboard visuals.
September 2026 Guided streaming-data monitoring (Generally Available) Guided streaming-data monitoring walks you through building no-code monitoring in Real-Time Hub. For more information, see Get started with Real-Time Hub.
September 2026 Warehouse operations skill sqldw-cli (Generally Available) The Warehouse operations skill sqldw-cli runs bounded, read-only diagnostics for warehouse and SQL analytics endpoint issues and returns evidence, recommended actions, and validation steps. For more information, see Diagnose warehouse workloads with Warehouse operations skill sqldw-cli.
September 2026 Capacity overage (Generally Available) Capacity overage is now generally available for Fabric F SKUs and is enabled by default when you create a capacity. It pays for eligible excess capacity usage to help prevent throttling, up to a configurable rolling 24-hour threshold. For more information, see Capacity overage in Microsoft Fabric and Enable capacity overage in Microsoft Fabric.
September 2026 Vector indexes and vector search in SQL database in Fabric (Generally Available) Vector indexes and vector search in SQL database in Fabric are now generally available. For more information, see CREATE VECTOR INDEX and Vector indexes and vector search general availability.
September 2026 SharePoint List mirroring (Generally available) SharePoint List mirroring is now generally available. You can continuously replicate SharePoint list data into OneLake and use it across Fabric workloads. For more information, see SharePoint List mirroring.
September 2026 dbt job in Fabric Data Factory (Generally Available) dbt job in Fabric Data Factory lets you author, schedule, and monitor dbt projects natively with serverless execution, integrated testing and documentation, and governance through Microsoft Entra ID and SQL security policies. For more information, see dbt job in Microsoft Fabric.
September 2026 Result set caching Result set caching works by persisting the final result sets for applicable SELECT T-SQL queries, bypassing complex compilation and data processing of the original query to return queries faster. To learn more, review the preview announcement at Result Set Caching for Microsoft Fabric Data Warehouse. Result set caching is enabled by default.
September 2026 SQL query editor enhancements (Generally Available) SQL query editor enhancements add a faster results grid, scalable object explorer and IntelliSense, autosave controls, bulk query management, .sql file import and export, and integrations with OneLake analytics and semantic models. For more information, see SQL query editor.
September 2026 Fabric networking communication policies Admin API (Generally Available) The List Networking Communication Policies Admin API provides a tenant-wide, paginated view of workspace inbound and outbound networking policies. For more information, see List Networking Communication Policies.
September 2026 ODBC to ADBC driver transition (Generally Available) ADBC drivers for supported Power BI and Fabric connectors are available for validation before they become the default and embedded ODBC drivers are disabled. For more information, see Transition from ODBC to ADBC drivers.
August 2026 Activator rules in Eventstream (Generally Available) Activator rules in Eventstream let you create, edit, delete, and open alert rules without leaving Eventstream. For more information, see Add a Fabric Activator destination to an eventstream.
August 2026 COPY INTO with workspace identity (Generally Available) COPY INTO with workspace identity lets Fabric Data Warehouse load approved data from OneLake or Azure Data Lake Storage Gen2 by using the workspace identity. For more information, see Ingest data into your warehouse using the COPY statement.
August 2026 Data-driven styling and map improvements (Generally Available) Data-driven styling and map improvements add value-based colors, data legends, marker rotation, traffic flow overlays, map view options, and improved layer settings to Fabric Maps. For more information, see Configure layer settings in Fabric Maps.
August 2026 Eventstream MQTT connector (Generally Available) The Eventstream MQTT connector connects MQTT brokers to Eventstream with TLS, mutual TLS, custom certificate authorities, and private network connectivity. For more information, see Add an MQTT source to an eventstream.
August 2026 Fabric Data Agent integration with Microsoft Copilot Studio (Generally Available) Fabric Data Agent integration with Microsoft Copilot Studio adds a Fabric data agent to a Copilot Studio agent as a Fabric IQ Data MCP tool while retaining source permissions.
August 2026 Metadata Sync support for Delta Checkpoint V2 (Generally Available) Metadata Sync support for Delta Checkpoint V2 lets both legacy and new Metadata Sync discover and synchronize Delta tables that use the Checkpoint V2 format. For more information, see SQL analytics endpoint metadata synchronization.
August 2026 Set alerts from Anomaly Detector (Generally Available) Set alerts directly from Anomaly Detector lets you create an alert from a published anomaly detector configuration without switching to Real-Time Hub. For more information, see Anomaly detection in Real-Time Intelligence.
August 2026 Mirroring for Google BigQuery (Generally Available) Mirroring in Fabric offers a simple, low-latency way to continuously replicate your Google BigQuery data into OneLake with no ETL pipelines, no data movement, and no proprietary formats. Once in Fabric, mirrored BigQuery data is queryable in near real time across Power BI Direct Lake, the SQL analytics endpoint, Data Engineering, Data Science, Real-Time Intelligence, and Copilot. For more information, see Mirrored databases from Google BigQuery and Tutorial: Set up mirroring for Google BigQuery.
August 2026 Migration Assistant to SQL database (Generally Available) Migration Assistant to SQL database is generally available and lets you validate a source DACPAC for compatibility before provisioning a SQL database in Fabric. For more information, see Fabric Migration Assistant for SQL database.
August 2026 Warehouse IDENTITY columns (Generally Available) IDENTITY columns are now generally available in Fabric Data Warehouse. You can use IDENTITY_INSERT and reseed operations for migration and controlled data loads. For more information, see IDENTITY columns and Use IDENTITY columns to create surrogate keys.
August 2026 Real-Time Dashboard KPI visualization (Generally Available) The KPI visualization in Real-Time Dashboards is now generally available, providing at-a-glance metrics with goals, status indicators, trends, and conditional formatting. For more information, see Customize Real-Time Dashboard visuals and Apply conditional formatting in Real-Time Dashboards.
August 2026 PowerTable billing personas (Generally Available) Billing personas for PowerTable in Fabric Planning are generally available, with role-based billing for Planners, Stakeholders, and Viewers. For more information, see Billing and usage for planning in Fabric.
August 2026 Workspace customer-managed key APIs (Generally Available) Workspace customer-managed key APIs are generally available, so administrators can automate workspace encryption-key assignment and removal. For more information, see Customer-managed keys for Fabric workspaces and Assign workspace encryption.
August 2026 Data source routing in Fabric data agents (Generally Available) Data source routing in Fabric data agents selects the most relevant lakehouse, warehouse, semantic model, or KQL database for each question by using schema metadata, source descriptions, example queries, and routing rules. For more information, see Configure a Fabric data agent and Provide example queries.
August 2026 Capacity Overview Events in Real-Time Hub (Generally Available) Capacity Overview Events in Real-Time Hub (Generally Available) stream capacity summary and state signals so you can monitor utilization and health, detect throttling or lifecycle changes, and trigger alerts or automated actions. For more information, see Introduction to Azure and Fabric events and Explore Fabric capacity overview events.
August 2026 Fabric Runtime 2.0 (Generally Available) Fabric Runtime 2.0 (Generally Available) includes Apache Spark 4.1, Delta Lake 4.2, Python 3.13, Java 21, Scala 2.13, and Azure Linux 3.0, and remains opt-in until its planned default rollout. For more information, see Runtime 2.0 in Fabric.
July 2026 Custom CA and mTLS in Eventstream connectors (Generally Available) Custom CA and mTLS support in Eventstream streaming connectors lets you use custom CA and client certificates from Azure Key Vault with the MQTT, Apache Kafka, AWS MSK, and Confluent Cloud for Apache Kafka source connectors. For more information, see Add and manage eventstream sources.
July 2026 Native execution engine UDFs and complex types (Generally Available) In Fabric Data Engineering, native execution engine support for Python UDFs, Scala UDFs, and complex data types accelerates Spark processing with no changes to existing notebooks or jobs. For more information, see Native execution engine UDF and complex type support.
July 2026 Materialized lake views analytics and insights (Generally Available) Materialized lake views analytics and insights show refresh trends, failure patterns, and prioritized recommendations so you can improve materialized lake view performance and cost. For more information, see View insights for materialized lake view runs.
July 2026 Usage-based resource estimations (Generally Available) Usage-based resource estimations help the Fabric Data Warehouse query optimizer use prior execution cardinalities to correct under- and over-estimates for more efficient resource requests and concurrency.
July 2026 Folders in Eventhouse tree (Generally Available) Folders in Eventhouse tree let you create, rename, delete, and move folders in the Eventhouse UI to organize tables, shortcuts, materialized views, functions, and data streams. For more information, see Manage and monitor a KQL database table.
July 2026 Eventhouse update policies with accelerated shortcuts (Generally Available) Eventhouse update policies with accelerated shortcuts let update policy queries join to accelerated shortcut external tables for governed ingestion-time enrichment.
July 2026 Fabric Maps tilesets (Generally Available) Fabric Maps tilesets create PMTiles from OneLake geospatial data so large static spatial datasets render as fast, interactive maps in Fabric. For more information, see Create tilesets.
July 2026 Real-Time Dashboard (Generally Available) Real-Time Dashboard (Generally Available) lets you monitor Eventhouse data with live refresh, Copilot-created tiles, drill-down exploration, Activator alerts, and team sharing. For more information, see What is Real-Time Dashboard? and Create a Real-Time Dashboard.
July 2026 Spark job monitoring (Generally Available) Job concurrency and queue monitoring (Generally Available) adds a capacity-level view so you can compare Spark resource use across workspaces, identify contention, and tune limits. For more information, see Job concurrency and queue monitoring and Concurrency limits and queueing in Apache Spark for Fabric.
July 2026 Resource Instance Rules for OneLake Network Security for Azure Services (Generally Available) Resource Instance Rules for OneLake let you allow specific Azure resource instances to access OneLake over the public endpoint while Fabric still evaluates data permissions. For more information, see Manage workspace inbound access with Resource Instance Rules.
July 2026 Lakehouse Query Explorer (Generally Available) Lakehouse Query Explorer lets you run Spark SQL in the Lakehouse explorer to inspect data, test joins, create views, and iterate before you move logic to notebooks or SQL analytics endpoints. For more information, see Query data with the Lakehouse query explorer.
July 2026 OneLake item-size reporting (Generally Available) OneLake item-size reporting gives you a workspace storage report that breaks down item data, system data, soft-deleted data, and billing status in the Fabric portal. For more information, see Get the size of OneLake items.
July 2026 OneLake storage tiers and lifecycle management (Generally Available) OneLake storage tiers and lifecycle management let you keep active data in the hot tier and move less-used data to cool or cold tiers with lifecycle policies. For more information, see Get the size of OneLake items and OneLake diagnostics.
July 2026 Plan (Generally Available) Plan in Fabric IQ (Generally Available) is a unified no-code platform for collaborative planning, reporting, analytics, data integration, and management. For more information, see What is plan?
July 2026 Dropped warehouse recovery (Generally Available) Dropped warehouse recovery lets Workspace Administrators restore accidentally dropped warehouses from the Workspace Recycle Bin within a configurable retention window (7 to 90 days), recovering table schemas, data, snapshots, permissions, views, and stored procedures. For more information, see Manage workspaces.

Community

This section summarizes new Microsoft Fabric community opportunities for prospective and current influencers and MVPs.

Power BI

Important

To access Power BI, use a web browser that's Chrome 94 or newer, Microsoft Edge 94 or newer, Safari 16.4 or newer, Firefox 93 or newer, or an equivalent version. If you use an older version, upgrade your web browser to a newer version by August 31, 2024. If you use an outdated browser version after this date, you might be prevented from accessing features in Power BI.

For updates to Power BI Desktop and the Power BI service, see What's new in Power BI?

Stay up to date

To stay up to date with the latest news and features, see the Microsoft Fabric Updates Blog.

For Microsoft Fabric release plans, see the Microsoft Fabric Roadmap.

Related content

  • Last updated on