Cube
📊 Cube Core is open-source semantic layer for AI, BI and embedded analytics
Last verified:
What is Cube?
Cube is a universal semantic layer platform that connects data sources to data apps, embedded analytics, BI tools, LLMs, and AI agents. It delivers consistent, secure, and scalable analytics by bringing clarity, consistency, and context to business data. Cube defines metrics upstream in a centralized semantic layer, eliminating duplicate data modeling efforts across different BI tools and data warehouse silos.
Key features include data modeling in a semantic layer IDE, Cube Store caching for sub-second query latency, support for all SQL-enabled data sources (Snowflake, BigQuery, Presto, Athena, Postgres), REST/GraphQL/SQL APIs for embedded analytics, row-level access control, pre-aggregations for performance optimization, and AI-powered features like semantic model agent, analytics chat, and Agentic IDE. Cube Core is the open-source headless version, while Cube Cloud is the fully managed enterprise-grade hosted solution.
Cube is designed for data engineers, developers, and data professionals who need to build modern data applications, embedded analytics, and BI solutions. It serves hobbyists completing small projects (Free tier), professionals streamlining data workflows (Starter tier), teams scaling analytics (Premium tier), and enterprises demanding security and compliance (Enterprise tier). The platform is trusted by companies like SecurityScorecard, Webflow, The Linux Foundation, Cloud Academy, and SamCart.
Cube pricing
Pricing model: Freemium
Cube offers four pricing tiers. Free tier is free forever for hobbyists with up to 5 workbooks, basic data source connections, 1,000 daily requests, and 1 day query history. Starter tier is $40 per Developer per month with unlimited workbooks, extended query limits (10K shared/50K dedicated), production deployment compute, Cube Store caching, Semantic Layer Sync, and observability. Premium tier is $80 per Developer per month with Explorer role ($40/user/month), Viewer role ($20/user/month), embedded dashboards, embedded analytics chat, unlimited queries, 99.950% uptime SLA, custom domains, multi-cluster deployment, and 7 days query history. Enterprise tier has custom pricing with 99.990% uptime SLA, dedicated single-tenant installation, BYOC, BYOLLM, SSO with SAML 2.0, workspace access control, DAX API for Power BI, and full audit trail. Self-serve customers pay monthly via Stripe credit card; order form customers pay annually.
Cube pros
- Open-source Cube Core available for self-hosting
- Universal semantic layer connects all data sources to all tools
- Supports all SQL-enabled data sources including Snowflake and BigQuery
- Cube Store caching provides sub-second query latency
- Built-in relational caching engine for high concurrency
- REST, GraphQL, and SQL APIs for embedded analytics
- Row-level access control for granular data security
- Pre-aggregations accelerate analytics performance
- AI-powered semantic model agent builds initial models
- Analytics chat enables natural language data queries
- Agentic IDE for building data applications
- Semantic Layer Sync keeps models up-to-date
- 99.950% uptime SLA on Premium tier
- 99.990% uptime SLA on Enterprise tier
- Bring Your Own Cloud (BYOC) on Enterprise tier
- Bring Your Own LLM (BYOLLM) on Enterprise tier
- SSO with SAML 2.0 for enterprise security
- Unlimited workbooks on Starter and higher tiers
- Unlimited queries on Premium and Enterprise tiers
- Custom domains supported on Premium tier
- Multi-cluster deployment on Premium tier
- Performance Insights available on Premium tier
- 18,000+ GitHub stars with active open-source community
- 13,000+ community members on Slack
Cube cons
- Free tier limited to 5 workbooks
- Free tier has 1,000 daily request limit
- Free tier only retains 1 day query history
- Starter tier lacks Premium LLMs access
- No ACH or wire transfer for self-serve plans
- Requires credit card via Stripe for self-serve
- Premium LLMs only available on Starter and higher
- Dedicated deployment costs extra at $0.60/hour
- Cube Store caching costs additional per hour
- Multi-cluster deployment costs $1.20/hour/cluster on Premium
- Learning curve for semantic layer data modeling
- Initial time investment needed to set up data models
- JavaScript data models required for dynamic features
- YAML data models lack some JavaScript functionality
- Enterprise pricing is custom (not publicly listed)
- Explorer role only available on Premium tier
- Viewer role only available on Premium tier
- Full audit trail only on Enterprise tier
- Advanced networking security only on Enterprise tier
- Cube Core lacks frontend and AI features
- Self-hosting requires your own infrastructure management
- Order form customers need annual upfront commitment
Frequently asked questions about Cube
What is Cube and what does it do?
Cube is a universal semantic layer that connects all your data sources with all your data apps, embedded analytics, BI tools, LLMs, or AI agents. It brings clarity, consistency, and context to business data by defining metrics upstream in a centralized semantic layer, eliminating duplicate data modeling across different tools. Cube delivers consistent, secure, and scalable analytics to every data consumer in the enterprise.
What is the difference between Cube Core and Cube Cloud?
Cube Core is the open-source headless semantic layer that you can self-host with Docker on your own infrastructure. It lacks frontend and AI features but can power your own embedded analytics or AI agents. Cube Cloud is the fully managed, enterprise-grade hosted solution with AI features like semantic model agent, analytics chat, Agentic IDE, workbooks, dashboards, plus enterprise features like SLAs, SSO, and dedicated installations.
What data sources does Cube support?
Cube was designed to work with all SQL-enabled data sources, including cloud data warehouses like Snowflake and Google BigQuery, query engines like Presto and Amazon Athena, and application databases like Postgres. You can connect any data source in the Free tier and above.
What is the Cube Store caching feature?
Cube Store caching is a built-in relational caching engine that provides sub-second latency and high concurrency for API requests. It accelerates query performance through pre-aggregations. This feature is available on Starter tier and higher, with pricing at $0.15/hour/worker for small caching on Starter and $0.30/hour/worker on Premium.
How does Cube's AI feature work?
Cube includes AI-powered features like semantic model agent that builds your initial semantic model automatically, workbooks and dashboards agent, and analytics chat for natural language data queries. The AI can generate SQL syntax, access control functions, and pre-aggregations. Cube's semantic layer provides business context to LLMs, increasing AI accuracy compared to text-to-SQL approaches.
What uptime SLAs does Cube offer?
Cube offers a 99.950% uptime SLA on the Premium tier and a 99.990% uptime SLA on the Enterprise tier. The Free and Starter tiers do not include an uptime SLA guarantee.
Can I use my own cloud provider with Cube?
Yes, the Enterprise tier includes Bring Your Own Cloud (BYOC) capability, allowing you to deploy Cube in your own AWS, GCP, or Azure infrastructure. This is an Enterprise-only feature for customers demanding maximum control over their deployment environment.
What authentication and security features are available?
Cube offers row-level access control on all tiers. Single Sign-On (SSO) with SAML 2.0 is available on Enterprise tier. Enterprise also includes workspace access control, full audit trail of all queries and actions, and advanced networking security features like VPC Peering and AWS PrivateLink. Enterprise Premier adds customer-managed encryption keys.
How is Cube pricing structured for self-serve customers?
Self-serve customers are billed monthly via Stripe credit card. The card is automatically charged at the start of each billing cycle. Invoices bill active seats in advance for the upcoming period. Seat count changes during a billed period are charged in arrears on a prorated basis on the next invoice. Cube does not support ACH or wire transfer payments for self-serve plans.
What BI tools and integrations does Cube support?
Cube integrates with all kinds of data visualization tools and BI platforms through its REST, GraphQL, and SQL APIs. It supports embedded dashboards and embedded analytics chat on Premium tier. Enterprise tier includes DAX API for Power BI integration and unlimited BI tool integrations. The platform is designed to work with any BI software that can connect via these APIs.