Skip to main content
πŸ“– The AI Tool Bible
Cube preview image
Cube logo

Cube

βœ“ Text source-checked Oct 24, 2026

Cube is an agentic analytics platform built on a semantic layer for AI-native business intelligence and embedded analytics.

Source-checked against the vendor's site Β· rebuilt from the vendor's own pagesProfile updated Link checked Pricing captured How we rate

FreemiumΒ· Free plan is free forever; Starter plan is $40 per developer per month; Premium plan is $80 per developer per monthRAGMulti-model (not source-checked)8.1 / 10

In short

Cube is an agentic analytics platform using a semantic layer for AI-native BI and embedded analytics. It connects AI agents to governed data via MCP and APIs.

Cube is an agentic analytics platform for business intelligence and embedded analytics, built on an open-source semantic layer. The product serves two primary use cases: internal business intelligence and embedded analytics. For internal business intelligence, data teams and business users explore data with workbooks, dashboards, and Analytics Chat, which provides natural-language analysis grounded in a governed data model. For embedded analytics, software companies ship the same surfaces and underlying APIs inside their own products to deliver customer-facing experiences.

The vendor says that Cube provides AI agents with governed, reliable business data through MCP and APIs so that every decision is grounded in trusted context. The vendor says that the semantic layer keeps metric definitions consistent across every surface of the platform. The vendor says that performance is maintained through pre-aggregation and caching, ensuring every chart resolves against semantic definitions. Users can interact with data from Claude, ChatGPT, Cursor, or a Slack agent via the MCP server.

Brex selected Cube over dbt Semantic Layer and LookML, citing the semantic layer as the factor that makes AI useful at scale. Drata uses Cube as a single source of truth for metric definitions, powering customer-facing dashboards and AI-driven quarterly business reviews.

The Free plan is available forever for hobbyists completing small projects and showcasing work. The Starter plan costs $40 per developer per month and is aimed at professionals streamlining data workflows. The Premium plan costs $80 per developer per month and is intended for teams scaling analytics and collaboration. The Enterprise plan has custom pricing for enterprises demanding security and support, requiring a demo request.

Cube Core is headless and does not ship a user interface, allowing users to build a custom analytics experience. Cube Core can be run locally or self-hosted using Docker.

From the vendor's pages

According to Cube: Cube is an agentic analytics platform for business intelligence and embedded analytics, Built on an open-source semantic layer. Its pages list this strength: The core semantic layer is open-source.

β€” The AI Tool Bible editorial team

Pros

  • βœ… The core semantic layer is open-source.
  • βœ… AI agents receive governed, reliable business data through MCP and APIs.
  • βœ… The platform achieves sub-second performance via pre-aggregation and caching.
  • βœ… The system is multi-tenant by construction, with governance flowing from the model to customer permissions.
  • βœ… The product integrates with Claude, ChatGPT, Cursor, and Slack via an MCP server.

Pricing (captured )

TierMonthlyPer seatKey features
FreeFreeYes
  • Connect any data source
  • Model in semantic layer IDE
  • Explore data in workbooks
  • Publish workbooks as dashboards
Starter$40Yes
  • Extended limits on agents
  • Use premium LLMs
  • Unlimited workbooks
  • Extended query limit
Premium$80Yes
  • Explorer role
  • Viewer role
  • Embedded dashboards
  • Embedded analytics chat
EnterpriseCustomβ€”
  • 99.990% Uptime SLA
  • Dedicated single tenant installation
  • Bring Your Own Cloud (BYOC)
  • Bring Your Own LLM (BYOLLM)

Prices change; we show the grid as captured on Nov 17, 2026. Check current pricing on the vendor's site β†’

Facts

Platforms
api (not source-checked)
API available
Yes (not source-checked)
Free trial
Yes (not source-checked)
Open source
Yes (license: NOASSERTION)
GitHub stars
21,220 (checked Nov 13, 2026) (not source-checked)
Last GitHub push
Nov 13, 2026 (not source-checked)
First commit
Sep 2018 (not source-checked)
Model used
Multi-model (not source-checked)
Vendor pages
Changelog
Hacker News
32 stories Β· top 109 points Β· last Nov 2, 2026 Β· checked Nov 10, 2026
Earliest Wayback capture
2019-03-07 (earliest archived capture of the site, not a launch date)

Where this tool sits

Use cases

internal business intelligenceembedded analyticsnatural-language analysis

Recent updates from Cube

Frequently asked

How much does Cube cost?
Cube offers a free plan for hobbyists. The Starter plan is $40 per developer per month, and the Premium plan is $80 per developer per month. Enterprise pricing is custom and requires a demo request.
What are the main use cases for Cube?
Cube serves internal business intelligence and embedded analytics. It supports workbooks, dashboards, and natural-language analysis for internal teams, while allowing software companies to ship analytics surfaces inside their own products.
How does Cube integrate with AI tools?
Cube provides governed business data to AI agents via MCP and APIs. Users can interact with data from Claude, ChatGPT, Cursor, or a Slack agent through the MCP server, ensuring decisions are grounded in trusted context.
Can I self-host Cube?
Yes, Cube Core is headless and can be run locally or self-hosted using Docker. It does not ship a user interface, allowing users to build a custom analytics experience.
How does Cube ensure data consistency?
The semantic layer keeps metric definitions consistent across every surface. Performance is maintained through pre-aggregation and caching, ensuring every chart resolves against semantic definitions for reliable business data.

Compare Cube with similar tools

All alternatives β†’

Explore related

Reviews