This QuickStart lists all the new and public beta features released, as well as bugs fixed in June 2026.
It is summary in nature, and you should refer to the specific Sigma documentation links provided for more information.
Public beta features will carry the section text "Beta".
All other features are considered released (GA or generally available).
Sigma actually has feature and bug fix releases weekly, and high-priority bug fixes on demand. We felt it was best to keep these QuickStarts to a summary of the previous month for your convenience.
New first Friday features QuickStarts will be published on the first Friday of each month, and will include information for the previous month.
For those wanting to see what Sigma is doing on each week, release notes are now also available on the Sigma Community site. There, you can opt in to receive notifications about future release notes in order to stay on top of everything new happening at Sigma. You can also subscribe to automated updates in any Slack channel using the Sigma Community release notes RSS feed.
For more information on how to subscribe to release note notifications, see About the release notes
For more information on Sigma's product release strategy, see Sigma product releases
If something is not working as you expect, here's how to contact Sigma support


Monitor token consumption, data sources, tool calls, and user feedback for AI features from a single admin dashboard.
WHY IT MATTERS:
Cost visibility and governance are blockers for rolling AI features out broadly. The AI usage dashboard gives admins one place to track token spend, which data sources AI is querying, which tools agents are calling, and where users are flagging quality issues — the inputs you need to manage AI as a budget line, not a black box.
For example, looking at the Overview page of the AI Usage dashboard for the current week we can see all the key metrics of interest and drill into anything that interests us:

For more information, see AI usage
The Sigma Assistant usage dashboard is deprecated and will be unavailable after September 15, 2026. Configure the AI usage dashboard instead.
For more information, see Configure the AI usage dashboard
For security reasons, if Run queries as recipient is enabled, either as an organization requirement or for specific exports, admins must add recipients to an allowlist in export settings before they can be listed as Cc or Bcc recipients.
For more information, see Manage export frequency and authentication settings
Deployed documents that include Open Sigma document actions now automatically deploy the referenced target documents and update the action references accordingly.
For more information, see What gets deployed to a tenant
The Administration > Usage > Users > User Detail table now includes a dedicated License type column that indicates the license tier associated with each user's account type.

Organizations using Sigma Tenants can now manage tenant organization attributes from the parent organization by creating and assigning attributes through the Tenants section of the Administration portal.
For more information, see Create and manage tenant organizations
Sigma deployment now includes Azure Australia (australiaeast) in New South Wales, providing lower latency and improved performance for Australian customers, with a disaster recovery region in australiasoutheast.
Sigma now uses GPT 5.4 instead of GPT 5.1 if your OpenAI or Azure OpenAI account supports it.

You can now create and use AI columns on Databricks connections. AI columns let you construct dynamic prompts that reference specific table columns, useful for tasks like enriching, summarizing, and classifying data.
For more information, see Create AI columns (Beta)

Build agentic solutions natively in Sigma with Sigma agents. Agents provide AI capabilities to a dashboard or application based on a predefined context of the data elements in a workbook.
WHY IT MATTERS:
Sigma agents bring agent-style AI into the place where the data, governance, and audit trail already live — no separate orchestration layer, no shuttling context between systems. Builders define what an agent can see and do inside the workbook itself, so the agent inherits Sigma's permissions, lineage, and version history by default.

For more information, see Sigma agents, Build Sigma agents, and Example agent implementations
There are also two QuickStarts on this topic:
Unlocking Insights from Unstructured Text with a Sigma Agent
Build Conversational AI Apps with Chat Elements and Snowflake Cortex

Use MCP servers for third-party services with Sigma agents by adding them to Sigma as MCP tools.
WHY IT MATTERS:
MCP tools let Sigma agents reach into the systems your team already uses — ticketing, CRM, custom internal APIs — without writing connector code. Open MCP support keeps Sigma compatible with the broader agent ecosystem rather than locking customers into a single AI stack.

For more information, see Configure MCP tools

You can now create AI columns to enrich your data using natural language prompts. AI columns let you construct dynamic prompts that reference specific table columns, and are useful for tasks like enriching, summarizing, and classifying data.
WHY IT MATTERS:
AI columns turn enrichment, classification, and summarization into a one-step workflow inside Sigma — no separate pipeline, no Python, no waiting on data engineering. Analysts can attach language model intelligence directly to existing tables, which is the fastest path from raw operational data to ready-to-analyze attributes.

For more information, see Create AI columns (Beta)
You can now apply the "kind":"mcp" tag to monitor queries sent by the Sigma MCP server — useful for tracking costs or usage associated with the MCP server.
For more information, see Monitor MCP queries

Sigma Assistant can now help design and build workbooks using natural language prompts. While editing a workbook, Assistant can explore any data source you have access to and assemble dashboards with charts, tables, KPIs, and filters, set up structured data entry, and scaffold AI-powered apps.
WHY IT MATTERS:
This shifts Assistant from answering questions to building the workbook with you — charts, tables, KPIs, filters, data-entry flows, and AI apps all generated from natural-language prompts against a permitted data source. For analysts and builders who already know what they want, it removes the click-by-click assembly step that takes the longest in dashboard work.

For more information, see Use AI to build dashboards and apps
Starting July 6, 2026, the documentation MCP server will provide only the searchDocs tool, which searches Sigma documentation and returns relevant content sections and source URLs.c

Integrate Snowflake Cortex Agents or Databricks Genie Spaces for use with Sigma Assistant and Sigma agent workflows.
WHY IT MATTERS:
This is the collaborative AI pattern — Sigma calls into Snowflake Cortex and Databricks Genie rather than competing with them. Customers keep their AI investments in the warehouse, while Sigma supplies the governed runtime, audit trail, and collaboration layer on top.
For more information, see Use warehouse agents with Sigma

Disable a sequence to prevent it from running, then enable it to run again. The disabled state persists across workbook refreshes and user sessions until you manually re-enable the sequence:

For more information, see Disable and enable sequences
The pause/resume option for action sequences has been deprecated and replaced by disable/enable controls. Because the pause state was session-based and not saved, no migration is needed. New and existing sequences automatically reflect the updated controls.

The Update a deployment policy endpoint (PATCH /v2/deploymentPolicies/{deploymentPolicyId}) is now available to update an existing deployment policy in Sigma.
For more information, see Update a deployment policy
The Sync a connection by path endpoint (POST /v2/connections/{connectionId}/sync) now supports syncing all connection objects by providing an empty path parameter.
For more information, see Sync a connection by path
The List files and List member files endpoints now include a parentSourceUrlId in the response when returning details about documents deployed to a tenant organization.
For more information, see List files and List member files

1: Collapsed and expanded row and column groupings now export as they appear in the workbook. Previously, all rows and columns exported fully expanded.
2: Tagging a workbook version that contained custom SQL failed with the error Cannot read properties of null (reading 'connectionId').
3: License tier usage counts in Administration > Users now align with the metrics shown on the Usage dashboard.
4: The value column position in pivot table exports now always matches how it appears in the workbook.
5: After making edits in a securely embedded workbook or a Sigma Public app and leaving the draft without publishing, edits were no longer visible in the draft.
6: After releasing warehouse agents for Sigma Assistant, asking questions of data source tables with Sigma Assistant was no longer available.
7: Editing a scheduled email or Slack export no longer displays an error that prevented the scheduled export from being edited.
8: Swapping sources with the v2/workbooks/swapSources endpoint with the copyInputTableData option set to True did not copy input table data successfully.
9: When editing or customizing elements in a container, tabbed container, or repeated container, you can now select the parent container from the context menu.
10: When modifying export schedules, only the owner of the schedule or an admin can make changes to the schedule.
11: Documents containing custom SQL elements failed to deploy when the source swap policy associated with the deployment policy defined specific paths in the connection to swap to or from.
12: Improved the loading time for KPI charts when opening a workbook for the first time.
13: When using warehouse agents with Sigma Assistant, the name and email address of the user is now shared with the agent.

Build Conversational AI Apps with Chat Elements and Snowflake Cortex
End-to-end pattern: a Sigma agent in a workbook calls a Snowflake Cortex Agent as a warehouse tool, with an optional input-table write-back flow.
Fundamentals 01: Overview (revised content)
The AI section is now split into dedicated sections (AI in Sigma, Sigma Assistant, Sigma agents, chat element, Formula Assistant and MCP Server).
Unlocking Insights from Unstructured Text with a Sigma Agent
Eight-step build walkthrough using a Sigma agent against Gong call transcripts in Snowflake. No code, no NLP pipeline.

You can add client and server certificates to Sigma, allowing you to configure mutual transport layer security (mTLS) for API connectors.
WHY IT MATTERS:
mTLS strengthens the trust chain between Sigma and customer-managed APIs, satisfying enterprise security review and compliance requirements where one-way TLS is not enough. This is a foundational capability for regulated industries and customers running Sigma against private, sensitive API endpoints.
For more information, see Configure mutual transport layer security for API connectors
You can now manually specify an OAuth provider when configuring OAuth as the single sign-on (SSO) method for authenticating to Sigma. This enables the use of custom domain names.
You can now select which Snowflake role to use when performing tasks in Sigma via OAuth connections.
For more information, see Choose a Snowflake role


Adding ad hoc calculated columns in pivot tables is now supported, enabling one-off calculations in a pivot table column without modifying the underlying dataset or restructuring the pivot.
An example: Calculating the average price of a product category
WHY IT MATTERS:
Ad hoc calcs are the fastest way to test a "what if I had this column?" idea without touching the underlying dataset or restructuring the pivot. Bringing them to pivot tables unblocks one of the most common asks in operational analytics — derive a metric, see it side-by-side with the existing rollups, and move on if it doesn't pan out.
For more information, see Create ad hoc calculations

Buttons and icon buttons are now a single button element with a new Button type selector for switching between the two styles. Icons remain available for use in navigation elements.
WHY IT MATTERS: Clickable icons are a core building block for app-style UX — compact action triggers that don't crowd the layout the way text buttons can. Merging buttons and icon buttons into one element gives builders a single, consistent way to add actions and more room to use icons for visual hierarchy.

The Unnest feature lets you create a table from an array of values. After selecting a data source and identifying a column with array data, Unnest creates a new table element with one row for each array item.
For more information, see Create a table from an array

You can enable custom page sidebars to appear on one or more workbook pages. Page sidebars allow you to repeat contents like filters or navigation options across multiple pages.
WHY IT MATTERS:
Custom page sidebars give builders a clean way to keep filters, navigation, or contextual content visible across multiple pages without rebuilding it each time. The result is faster workbook authoring and a more consistent reading experience for end users.
Here is a short demo (there is no audio on this one):
For more information, see Add custom page panels to a workbook

Hierarchy columns are now generally available.

WHY IT MATTERS:
Drill paths and parent/child rollups are core to operational reporting — finance hierarchies, org charts, product taxonomies. Bringing hierarchy columns to GA gives every builder a stable, supported way to model these structures directly in Sigma instead of pre-flattening them upstream.
For more information, see Work with hierarchies and RaggedHierarchy
You can now configure text color options on navigation elements for finer styling control.
For more information, see Use the navigation element to guide user exploration

Build a progress bar or progress ring to display a value as a percentage or proportion of a target value.
WHY IT MATTERS:
Goal and target tracking shows up everywhere — pipeline coverage, OKRs, fundraising, capacity. Progress bars and rings replace the usual "calc a percentage and explain it" pattern with a clear visual that immediately communicates how close you are to a target.
For more information, see Build a progress bar and Build a progress ring
When segmented control values are too long to display fully when viewing or editing a document, the control converts to a drop-down menu instead of truncating the text.
For more information, see Segmented control
When tagging a workbook version, you can swap the API connectors used by Call API actions and the MCP tools and warehouse agents used by Sigma agents.
For more information, see Tag a document version
Several UI strings have shifted from first-person to second-person to make ownership clearer.
My workspaces is now Your workspaces
My last activity column in the Recent view is now Your last activity
The saved-view options Set as my default view and Remove as my default view are now Set as your default view and Remove as your default view.
Embed customers should review any custom UI references to these strings.
You can now use dynamic text when formatting the Subject and Message fields in ad hoc and scheduled email exports.
For more information, see Export to email
The #raw directive in custom SQL is deprecated. On December 1, 2026, Sigma will no longer support the #raw directive. Update any custom SQL that uses #raw to use the #identifier directive before that date to avoid query errors.
For more information, see Replace the #raw directive in custom SQL

Additional Resource Links
Blog
Community
Help Center
QuickStarts
