Pause Auto-Runs in Explore and Workbooks

Build your query first, then execute, or set after_run: false on a view by default.

Brand All of Cube Cloud

Light and dark themes, colors, logos, and fonts — applied across the app and embedded surfaces with live preview.

Encode business definitions, calculation methods, and domain terminology as rules the agent follows when answering or analyzing.

Define reusable multi-step workflows the Cube agent can run — analyze, build reports and dashboards, then summarize the findings.

A chat panel docked on any published or embedded dashboard — ask questions in plain language, answered with live queries against your model.

Cube's MCP server now exposes multiple Cube agents, letting the MCP client pick the right one based on each question's context.

Let the Cube AI agent run real Python for forecasting, statistical tests, and modeling — going beyond what SQL aggregation alone can do.

Set the dashboards, workbooks, and folders embedded users see on first login, so they start with useful content instead of an empty workspace.

Map charts now support regions, plus a round of refinements — visualize geographic data with more control inside workbooks.

Query your semantic model directly in SQL — reference measures and dimensions by name instead of rebuilding the logic in raw SQL.

Audit every access policy in your data model — row-level filters, member restrictions, and masking — broken down by user group.

Share an Analytics Chat thread with teammates as a read-only conversation — per-user, per-group, or org-wide access.

Creator Mode users can now share workbooks with teammates from inside your embedded app — per-user or group access.

A deep dive into Cube's Table chart — column controls, conditional formatting, text and background styling, sorting, and pivots.

Organize semantic views into named collections by domain — helping agents, embedded apps, and people find the right dataset.