Cube pricing guide details
Pricing Overview
Cube publishes a price list. The Free plan is free forever for small projects. Paid plans price seats by role: a Developer seat costs $40 per user per month on Starter and $80 on Premium, and Premium adds Explorer seats at $40 and Viewer seats at $20 per user per month. Infrastructure is billed by the hour on top of seats, and Enterprise is custom. The open-source Cube Core remains free, which is a genuine advantage for teams with the engineering resources to self-host.
Plan Comparison
Cube Cloud lists four plans: Free, Starter, Premium and Enterprise. The open-source Cube Core can also be self-hosted.
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Feature | Free | Starter | Premium | Enterprise |
|---|---|---|---|---|
| Developer seat | $0 | $40 / user / month | $80 / user / month | Custom |
| Explorer and Viewer seats | — | — | $40 and $20 / user / month | Custom |
| Daily request limit | 1,000 | 10K shared, 50K dedicated | Unlimited | Unlimited |
| Workbooks | Up to 5 | Unlimited | Unlimited | Unlimited |
| Dedicated deployment | — | $0.60 / hour | $1.20 / hour | Custom |
| Query history retention | 1 day | 1 day | 7 days | Custom |
| Embedded dashboards and analytics chat | — | — | Yes | Yes |
| Uptime SLA | — | — | 99.950% | 99.990% |
| SSO, custom roles, audit logging | — | — | — | Yes |
| Support | Community | Community; limited chat and email | Community; chat and email | Community; chat and email; dedicated CSM |
The free self-hosted option gives you the full semantic layer engine, but you handle infrastructure, upgrades, and scaling yourself. Cube Cloud's free tier lets you experiment without commitment, but production workloads will almost certainly require a paid plan. Enterprise pricing is negotiated individually; on the published plans, the bill is seats by role plus infrastructure at $0.60 to $1.20 per hour for a dedicated deployment.
Hidden Costs and Considerations
Self-hosting Cube Core looks free on paper, but infrastructure costs for compute, caching layers, and database connections add up quickly. On Cube Cloud, hourly infrastructure charges for the dedicated deployment, additional API instances and Cube Store caching come on top of seats, so the bill grows with the compute you run. We also flag that enterprise features like SSO, advanced RBAC, and dedicated support are gated behind custom contracts. Budget for onboarding time as well, since building a semantic layer requires significant upfront data modeling effort regardless of which plan you choose.
How Cube Pricing Compares
Cube sits in a unique position as a semantic layer platform competing against broader BI and analytics tools. Here is how it stacks up against common alternatives on cost.
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Tool | Free Tier | Starting Paid Price | Pricing Model | Best For |
|---|---|---|---|---|
| Cube | Yes (open-source + cloud) | $40/developer/month (Starter) | Seats by role plus hourly infrastructure | Semantic layer, embedded analytics |
| Amazon QuickSight | Yes (5 users) | $3/user/mo for Readers and $24 for Authors | Per-user | AWS-native BI dashboards |
| KNIME | Yes (open source) | €19/mo | Tiered subscription | Data science workflows |
| Amplitude | Yes | From $0, scales with events | Freemium | Product analytics |
Cube's pricing is harder to compare directly because it combines seats priced by role with hourly infrastructure charges. For teams that need a semantic layer to power multiple downstream BI tools and AI agents, Cube can be more cost-effective than licensing separate BI platforms. However, if you simply need dashboards and reports, Amazon QuickSight at $3/user/mo for Readers and $24 for Authors offers a more predictable cost structure. KNIME serves a different use case entirely, focusing on data science rather than analytics serving. We recommend Cube primarily for organizations that treat their semantic layer as core infrastructure rather than as a reporting add-on.