Skip to content
Roadmap

What's GA, what's coming.

Every Databricks product and feature we track, with the status it was last seen in. Statuses come from Databricks docs release notes, or from a public Databricks announcement where no note exists yet. A claim from an announcement is checked by hand before it changes a status. Independently maintained, not affiliated with Databricks. Last updated today.

180 tracked·108 GA·59 in preview·11 announced

Platform

Apps, marketplace, foundations.

5
  • Custom URLsPublic previewGives a Databricks account a single branded domain, such as mycompany.databricks.com, that serves every workspace in the account, so users log in once and stay signed in as they move between workspaces. Existing per-workspace URLs keep working, and admins can redirect them to the custom URL.
  • Databricks Academy ProGAAn annual per-seat training subscription that bundles unlimited self-paced courses, live instructor-led classes and hands-on labs in a real Databricks environment. Each seat also includes one certification exam voucher per year.
  • Databricks AppsGABuilds and hosts data and AI applications written in Python, for example with Streamlit, Dash or Gradio, or in Node.js, on serverless compute inside a Databricks workspace. Apps integrate with Unity Catalog for governance, Databricks SQL for queries and OAuth for authentication, and are used for dashboards, RAG chat apps, data entry forms and operational tools.6
  • Databricks IDE ExtensionsGAThe Databricks IDE extension, also known as the Databricks extension for Visual Studio Code, connects Visual Studio Code or Cursor on a local machine to remote Databricks workspaces. Developers can sync local code with the workspace, run Python files and notebooks on Databricks compute or as Lakeflow Jobs, debug with Databricks Connect, and deploy Declarative Automation Bundles.
  • Databricks MarketplaceGADatabricks Marketplace is a platform for discovering and deploying data products and solutions.

Governance

Unity Catalog, ABAC, classification, audit.

1

Compute

Clusters, serverless, runtimes.

4
  • DataFrame caching on serverless computePublic previewLets df.cache() and df.persist() work on serverless compute: Databricks computes the DataFrame once, materializes the result as a snapshot and reuses it for the rest of the session. It has been in Public Preview since October 23, 2026, on by default for most workspaces but not available with the compliance security profile, and needs serverless environment version 5 or above, Databricks Connect for Python 18.1.8 or above or Databricks Connect for Scala 18.1.5 or above.
  • Lakehouse Real-TimeBetaLakehouse Real-Time (Lakehouse//RT) is a serverless SQL warehouse type for low-latency, high-concurrency workloads: sub-second SQL read queries on Unity Catalog managed tables in Delta Lake or Apache Iceberg format, materialized views, streaming tables and metric views, for serving analytical data to applications, operational analytics and dashboards with hundreds to thousands of concurrent users. It is created and permissioned like other SQL warehouses and accepts connections only through the Statement Execution API. In Beta since June 30, 2026: the Databricks account team enables it for an account and a workspace admin then turns on the Lakehouse RT preview. AI/BI dashboards work with it since October 15, 2026, and since November 3, 2026 it reads tables protected by row filters, column masks and ABAC policies.
  • Serverless compute access controlGALets workspace admins control who can run serverless workloads through two built-in serverless compute objects: Default Interactive Compute for notebooks and Databricks Connect, and Default Automated Compute for jobs and Spark Declarative Pipelines. Each object has two permission levels, Can Use and Can Manage, and all workspace users have Can Use by default. Databricks SQL, Model Serving, Lakebase and Databricks Apps are not covered by these controls.1
  • Shared ServerlessBetaLets notebooks and files in the same Git folder or Declarative Automation Bundle share one serverless compute session and one Python environment, defined by an optional pyproject.toml file at the folder or bundle root (serverless environment version 5 or above). Git folders got it in Beta on September 2, 2026 and bundles on October 22, 2026, when serverless notebook and Python script job tasks also began taking their environment from pyproject.toml (Beta); workspace admins turn it on with the Project-Scoped Shared Compute preview.

Storage

Delta Lake, Lakebase, table formats.

5

Data engineering

Lakeflow, pipelines, ingestion.

4
  • AUTO CDCGAThe AUTO CDC and AUTO CDC FROM SNAPSHOT APIs in Lakeflow pipelines, which process a change data feed or a series of database snapshots into SCD Type 1 or Type 2 tables without hand-written merge logic. They replaced the APPLY CHANGES APIs, with the same syntax, and the FLOW AUTO CDC clause brings the same flows to streaming tables in Databricks SQL. Open-source Spark Declarative Pipelines has its own Auto CDC, separate from these APIs: SCD Type 1 arrived in Apache Spark 4.2 and SCD Type 2 in Apache Spark 4.3 (October 2026).1
  • LakeflowGADatabricks' end-to-end data engineering solution for ingestion, transformation and orchestration. It includes Lakeflow Connect for ingestion connectors, Lakeflow pipelines built on Spark Declarative Pipelines (formerly DLT), Lakeflow Jobs (formerly Databricks Jobs) for orchestration, and Lakeflow Designer, a visual tool for building transformations with a drag-and-drop canvas or natural language. Since October 26, 2026 Lakeflow Jobs show query performance metrics and insights at the job level for supported serverless tasks, with an Optimize button that opens Genie Code with the insights (generally available; on by default for eligible workspaces with the compliance security profile on or after November 25, 2026). Since November 3, 2026 a job can also be triggered when one to ten upstream jobs complete, including jobs in other workspaces on the same Unity Catalog metastore, with the dependency configured on the downstream job alone (Beta, turned on from the Previews page in the downstream and every upstream workspace).3
  • Real-time mode in Spark Structured StreamingGAA Structured Streaming trigger that processes each event as it arrives instead of in micro-batches, giving end-to-end latency in milliseconds for workloads such as fraud detection and personalization. It runs on classic compute with dedicated or standard access and supports stateful work: aggregations over tumbling and sliding windows, deduplication, joins and transformWithState. Real-time mode in Lakeflow pipelines is a separate row, generally available since October 2026, and open-source Apache Spark 4.3 (October 2026) brought stateful processing to the upstream Real-Time Mode.1
  • Zerobus IngestGAFully managed, serverless push-based streaming ingest API that writes data directly from any producer into Delta tables governed by Unity Catalog. Writing into streaming tables went generally available on August 26, 2026 and Apache Arrow support on September 1, 2026; writing into managed tables backed by default storage was in Public Preview from September 7, 2026 and has been generally available since November 2, 2026, though some workspaces, including those with the compliance security profile, still turn it on from the Previews page. The C++ and C#/.NET SDKs are in Beta.

AI & ML

Agent Bricks, Genie, MLflow, AI Search.

22
  • Agent BricksGAThe Databricks agent developer platform for building, deploying and governing AI agents in code, with any framework and any model. Its first no-code agents, Knowledge Assistant and Supervisor Agent, reached GA in early 2026 and became legacy in October 2026. Newer parts such as the Agent Bricks CLI, Databricks Sandbox and managed agent memory are in Beta.3
  • AI Functions REST APIGAA REST API for calling ai_parse_document, ai_extract and ai_classify in real time from applications and agents, for example as soon as a document arrives. Databricks also markets it as the Document Intelligence real-time REST APIs.
  • AI Gateway (legacy) for serving endpointsDeprecatedThe original Mosaic AI Gateway for model serving endpoints, which adds permission and rate limiting, payload logging, usage tracking, AI guardrails, fallbacks and traffic splitting to individual endpoints. Databricks now calls it AI Gateway (legacy) and recommends the newer Unity Gateway instead.
  • AI RuntimeGAServerless NVIDIA GPU compute inside Databricks for training and fine-tuning models next to governed data, with GPUs that start in seconds and no GPU infrastructure to manage. It was first released as serverless GPU compute (Beta, June 2025), was named AI Runtime when it entered Public Preview in March 2026, and Databricks announced general availability on October 21, 2026. Since August 27, 2026, Genie Code also works in notebooks connected to AI Runtime, where it generates distributed training code, resolves environment and dependency issues, investigates GPU and distributed workload failures, and migrates Foundation Model Fine-tuning API and classic Databricks Runtime ML GPU notebooks; the release note placed this in Public Preview and no later note has moved it, although the docs page shows no preview label today.
  • AI Search (formerly Vector Search)GABuilds search indexes from Unity Catalog tables (Delta tables, streaming tables, materialized views and managed Iceberg v3 tables) and serves vector similarity, hybrid and full-text keyword queries from AI Search endpoints through a REST API or Python SDK; full-text search is in Beta. Delta Sync indexes stay in sync with their source table, with embeddings computed by Databricks or supplied by the user, while Direct Vector Access indexes are updated through the API. Standard endpoints hold up to 320 million vectors at dimension 768, and storage-optimized endpoints over one billion.7
  • ai_decideBetaA SQL AI function that answers one or more questions about text or structured data with a probability, a choice from named criteria, or a score on a scale. It is built for routing, triage and prioritization over governed data and can also be called through a REST API for real-time apps and agents. Since October 27, 2026 it can also be used in Lakeflow Designer's AI Function operator (Beta).
  • ai_enrichBetaA SQL AI function that generates new columns for each row from a schema you define, filling them from the input content and, when asked, grounding them in knowledge sources such as AI Search indexes, web search or a search tool on an external MCP server registered in Unity Gateway. It is in Beta behind the Ai Enrich workspace preview; October 20, 2026, when MCP grounding arrived (Beta), is the first dated release note that names it with its tier.
  • ai_extract precision modeGAA mode of the ai_extract function for complex document extraction, such as multi-page documents, schemas with 50 or more fields, many extracted items per document and conditional field logic. It pairs custom fine-tuned extraction models with an agentic harness that extracts in parallel with subagents and merges the results, and it is turned on with the precision mode setting in ai_extract or a toggle in the Information Extraction UI. It was previously listed here as Document Intelligence Precision Mode, after the title of the launch blog.
  • Claude Fable 5 on DatabricksGAAnthropic's Claude Fable 5 model, offered as a Databricks-hosted model through Foundation Model APIs pay-per-token on AWS, Azure and Google Cloud, with access governed through Unity Gateway.
  • CustomerLakeBetaDatabricks' agentic customer data platform (CDP) that runs inside Databricks under Unity Catalog governance, giving marketers Customer 360 profiles, natural language audience building with Genie, Campaign Agents and activation to martech and adtech tools without copying data out. It launched in Private Preview at Data + AI Summit on June 16, 2026 and reached Beta on October 19, 2026 when Campaign Agents moved to Beta, while Profile Agents are in Private Preview; access is through the Databricks account team.1
  • Databricks Feature StoreGAThe Databricks central registry for the features used in AI and ML models, governed in Unity Catalog with lineage, point-in-time joins, and feature sharing and discovery across the organization. Teams author features as Feature Views or feature tables and can publish them to Online Feature Stores, powered by Lakebase, for real-time serving. The older Workspace Feature Store for workspaces without Unity Catalog is deprecated.1
  • Databricks GenieGAThe Databricks family of AI experiences for working with data in natural language. It includes Genie One (formerly Databricks One) for business users, Genie Agents (formerly Genie spaces) that data teams curate, Genie Code (formerly Databricks Assistant) for developers, and the Databricks Genie mobile, Slack and Microsoft Teams apps. Genie spaces became generally available in December 2024, while the mobile, Slack and Teams apps are still in Public Preview.9
  • Foundation Model APIs provisioned throughputGAA Foundation Model APIs serving mode that runs foundation models on dedicated capacity with performance guarantees, and the mode Databricks recommends for production workloads. It serves base models, fine-tuned variants and fully custom weights of supported model architectures, with Databricks managing the serving infrastructure.
  • Genesis Workbench (solution accelerator)GAAn open-source Databricks solution accelerator for drug discovery and genomics that runs NVIDIA BioNeMo, Parabricks and other biology models behind a no-code app inside your own workspace. Anyone can deploy it from GitHub, but it is a Databricks Industry Solutions blueprint, not a platform feature with a preview or GA label.
  • Genie OntologyPublic previewThe context layer that gives Genie One and Genie Code a shared map of an organization's business knowledge. It combines Unity Catalog semantics, such as metric views, with snippets that Genie extracts and ranks from existing assets such as dashboards, SQL queries and Genie Agents, and it uses only snippets from assets the user has permission to see.1
  • MLflowGAMLflow is an open-source platform for managing the end-to-end machine learning lifecycle.2
  • Mosaic AI (retired brand)DeprecatedFormer umbrella brand for Databricks AI and ML tools such as Model Serving, Vector Search, Agent Framework and Agent Evaluation. Databricks has dropped the name: Agent Bricks became the umbrella for all its AI capabilities in March 2026, and the tools continue under plain names.
  • OmnigentBetaOpen-source meta-harness for composing, controlling, and sharing multiple AI agents and agent harnesses, also offered as a managed Databricks version.2
  • OpenAI GPT-5.5GAOpenAI's GPT-5.5 and GPT-5.5 Pro models are available as Databricks-hosted models in Model Serving, called through Foundation Model APIs pay-per-token.
  • PipelineIQ (Databricks internal tool)AnnouncedPipelineIQ is a sales pipeline tool that Databricks built for its own field sales organization on Foundation Model APIs, Unity Catalog, Delta Lake and AI/BI dashboards. It gives each sales opportunity a Walk, Pivot or Accelerate recommendation with a rationale and an action plan. Databricks shared it in a blog as an example of using its own platform; it is not a product or feature that customers can turn on.
  • Predictive AI FunctionsBetaA family of SQL functions for forecasting and analysis: ai_forecast version 2 projects time series forward with a time series foundation model, ai_top_drivers ranks the dimension values that contribute most to a change in a metric, and ai_detect_anomalies, ai_predict_value and ai_predict_class detect anomalies and predict values and classes. A workspace admin turns them on with the Predictive AI Functions option on the Previews page. Genie can call ai_top_drivers, and Databricks says the other functions are coming to Genie.
  • WASM Postgres for AI Agent SandboxesAnnouncedEmbeds lightweight WASM-based Postgres databases (PGlite) directly inside AI agent sandboxes with real-time synchronization back to Lakebase.

SQL & BI

Databricks SQL, dashboards, BI.

9
  • AI/BI MapsGAAI/BI dashboards render maps directly from Geometry or Geography columns.
  • Dashboard relationshipsGADashboard relationships model the join logic between the datasets of an AI/BI dashboard: an author defines a join once with its cardinality and the query engine resolves the joins each visualization needs at query time, which allows multi-fact, multi-grain data models without pre-joining in SQL or fan-out. Private Preview from June 18, 2026, Public Preview from July 1, 2026 and generally available since October 29, 2026; since July 30, 2026 Power BI and Tableau imports through Genie Code create the relationships they detect, and relationship-scoped filters, where a widget responds only to filters on related datasets, are in Public Preview since September 10, 2026.
  • Dataset materialization for AI/BI dashboardsBetaPre-computes AI/BI dashboard datasets on a schedule so published dashboards render from stored results instead of rerunning expensive queries on every load or click. It covers SQL datasets and local metric views and is turned on from the Previews page.
  • Full-text search indexesBetaA SQL-created index that accelerates substring and keyword queries on text columns of Unity Catalog managed Delta and Iceberg tables by tokenizing content and skipping non-matching files.
  • IP functionsGABuilt-in SQL, PySpark and Scala functions that parse, validate and canonicalize IPv4 and IPv6 addresses and CIDR blocks, such as ip_host, ip_network and ip_cidr_contains, with try_ variants that return NULL on invalid input. ip_cidr_contains also drives an optimized range join for matching addresses against CIDR blocks. The functions require Databricks Runtime 18.2 and above.
  • MATCH_RECOGNIZEBetaA SQL:2016 clause that finds row patterns in ordered data, such as runs of consecutive values or V-shaped price moves, and returns one row per match or every matching row. It works in Databricks SQL and Databricks Runtime 19.0 and above once a workspace admin turns on the Match Recognize preview.
  • NEAREST BY joinGAA SQL join that returns, for each input row, the top K closest rows from a target table ranked by a distance or similarity expression such as vector_cosine_similarity. It supports exact and approximate search for batch vector search on lakehouse tables in Databricks SQL and Databricks Runtime 18 and above.
  • Spatial SQLGANative SQL support for geospatial vector data on Databricks: the GEOMETRY and GEOGRAPHY column types, more than 90 OGC-compliant ST functions for importing, exporting, measuring, constructing and transforming shapes, and spatial joins. GEOMETRY and the ST functions are generally available, while GEOGRAPHY support is still in Public Preview.1
  • SQL scriptingGALets you write procedural logic in SQL, based on the ANSI SQL/PSM standard, with conditional statements, loops, local variables and exception handling. It helps move procedural code from legacy data warehouses together with related features that shipped separately, such as stored procedures, cursors and multi-statement transactions.

Connectivity

Lakeflow Connect, sharing, partner integrations.

5
  • Databricks Clean RoomsGADatabricks Clean Rooms uses OpenSharing and serverless compute to give up to ten parties a privacy-protecting environment where approved notebooks and JAR workloads run on combined data without any party seeing another's data. Public Preview from August 6, 2024, opened to all AWS and Azure customers on November 13, 2024, and generally available since January 29, 2025; in 2026 packaged clean rooms (July 31), shared output, JAR workloads and one-step identity resolution providers (October 13) reached GA, and since October 28, 2026 clean room notebooks can run on serverless GPU compute on AI Runtime for inference and training.
  • Databricks Excel Add-inGAA Microsoft Excel add-in that browses and imports governed Unity Catalog tables and views, runs SQL, builds pivot tables and writes data back to Unity Catalog. It works in Excel for the web and the Windows and macOS desktop apps.
  • Databricks JDBC DriverGAThe open-source Databricks JDBC Driver, version 3 and above, connects Java applications and tools such as DataGrip, DBeaver and SQL Workbench/J to Databricks. It supports OAuth, Cloud Fetch and Unity Catalog volume operations, and its source code is public under the Apache 2.0 license. The older driver below version 3 is now called Simba JDBC Driver (Legacy).1
  • MCP servers in Databricks MarketplacePublic previewLets providers list Model Context Protocol (MCP) servers in Databricks Marketplace, and lets Databricks users install those servers to connect AI agents, for example in Supervisor Agent and AI Playground, to external data sources, SaaS tools and developer APIs. It was previously listed here as MCP Marketplace.
  • OpenSharing on Databricks (formerly Delta Sharing)GAThe data sharing platform built into Databricks, called Delta Sharing until June 2026. It shares tables (managed Iceberg tables too, generally available since 2026-08-14), views, materialized views and streaming tables with other organizations whether or not they use Databricks, shares volumes, models and notebooks with recipients on Databricks, and is the basis for Databricks Marketplace and Clean Rooms. Recipients can apply their own row filters and column masks to shared tables (generally available since 2025-10-06), and sharing tables and schemas secured by ABAC policies entered Public Preview on 2025-11-17. Since 2026-06-15 Databricks-to-Databricks sharing also reaches other regulatory domains such as AWS GovCloud and Azure China (Public Preview), and since 2026-09-22 providers can invite a recipient by email address instead of exchanging a credential file (Beta).6

Observability

Monitoring, system tables, telemetry.

4
  • BudgetsGASpend monitoring in the Databricks account console. Account admins, and workspace admins for the workspaces they manage, create a budget for the whole account or filtered by workspace, product and custom tags, with up to four monthly thresholds in US dollars at list price that email chosen addresses when spend crosses them. Budgets scoped to Unity Gateway, which is also how Genie budgets work, add per-user limits, can block further requests at a threshold and update in near real time, while other budgets can lag up to 24 hours. Budgets went into Public Preview in June 2024 and reached GA in July 2026, and October 2026 added Budget Manager and Budget Viewer roles and a Spend Status API for Unity Gateway budgets. Since November 1, 2026, Unity Gateway budgets also count spend from task-specific AI Functions such as ai_parse_document, ai_extract and ai_classify, across batch and real-time execution, although those requests are not recorded in the system.ai_gateway.usage table.
  • Endpoint telemetry for custom model serving endpointsGAPersists OpenTelemetry logs, traces and metrics from custom model serving endpoints (and agent serving endpoints) to Unity Catalog tables, so teams can run root cause analysis, monitor endpoint health and keep records for compliance with standard SQL queries. It is configured per endpoint and needs a Unity Catalog-enabled workspace. In Beta from March 6, 2026 and generally available since June 11, 2026.
  • Hydra (Databricks internal tool)AnnouncedDatabricks' in-house platform for storing high-cardinality troubleshooting metrics on the lakehouse, described in a May 2026 engineering blog. Databricks engineers use it to debug the Databricks platform, and it is not a product that customers can use or turn on.
  • Query tags for SQL warehousesGAEnables tagging queries with custom metadata to facilitate granular compute cost attribution and tracking via system query history tables.

Security

Encryption, network, compliance.

4
  • Inbound Private Link for account-level resourcesBetaLets users on private networks reach Databricks account-level resources, such as the account console and account-level Genie One, through a Private Link endpoint instead of the public internet. This is separate from workspace-level inbound Private Link, which is an older, generally available feature.
  • LakewatchPrivate previewLakewatch is an agentic SIEM (security information and event management) product from Databricks. It brings security, IT and business data into one governed environment in open formats, and uses AI agents built with Agent Bricks, along with Genie, to automate threat detection, triage and investigation.
  • OAuth custom identity claimsGALets an application add a custom identity claim to the OAuth token it requests for a service principal and filter rows by that claim in a dynamic view or an ABAC policy with current_oauth_custom_identity_claim(), so one service principal can serve many end users who each see only their own data, with no Databricks user per end user. The claim reaches SQL through JDBC, the SQL Statement Execution API and the Genie Agents API, on a SQL warehouse or Databricks Runtime 19 or above. Released on October 19, 2026 with no preview label.
  • Private network gatewayPublic previewA managed gateway that connects Databricks serverless compute to resources in your own network, on-premises systems and the internet through one connection, so you can route serverless egress through your own security appliances or stable IP addresses. It is in Public Preview on Azure and Private Preview on AWS, with Google Cloud support in development.

How this list is maintained

Every entry links to a public Databricks source. Each night we read the Databricks blog and the docs release-note feed. A status moves forward on its own only when a release note states it; a claim from a blog post waits for a check by hand before it changes anything. We do not see the private roadmap Databricks shares with customers, so a feature that has not been mentioned in public is not here. See a stale entry or a missing one? Tell us. See what changed.