Cube and Mode Analytics were not competing tools — they solved fundamentally different problems. Cube is a semantic layer and metrics API that sits between your data warehouse and your analytics tools. Mode was a SQL-first analytics platform for data analysts doing exploratory work, now absorbed into ThoughtSpot Analyst Studio. Teams comparing them are likely conflating two different layers of the modern data stack. If you need a governed metrics layer, use Cube. If you need a SQL notebook for analyst work, ThoughtSpot Analyst Studio is Mode's successor.
Key Takeaways
| Category | Cube (2026) | Mode / ThoughtSpot Analyst Studio (2026) |
|---|---|---|
| Product Category | Semantic layer / metrics API | SQL analytics platform (now ThoughtSpot) |
| Availability | Active — open source + Cloud | Mode discontinued; ThoughtSpot Analyst Studio active |
| Primary User | Data engineers, backend engineers | Data analysts, SQL practitioners |
| Native Dashboards | No — powers other tools | Yes — reporting and notebooks |
| Open Source | Yes — cube.js on GitHub | No |
| Pricing Model | Consumption-based (CCUs) | Enterprise — ThoughtSpot pricing |
| Starting Cost | Free (self-hosted) / ~$3K-8K/mo cloud | ~$1,250+/mo (ThoughtSpot) |
| Best For | Building governed metrics infrastructure | SQL-based exploratory analytics |
What Is Cube?
Cube (formerly Cube.js) is an open-source semantic layer framework and managed cloud platform. Founded in 2019, Cube operates at a specific and often misunderstood position in the modern data stack: it is not a BI tool, not a dashboard builder, and not a query editor. It is the layer between your data warehouse and every analytics interface that consumes data.
The core concept of Cube is that business logic — how you define revenue, what counts as an active user, how you calculate margin — should be defined in one place and consumed everywhere. Cube enables this through a YAML- and JavaScript-based data model where you define cubes (think: tables or views), measures (metrics), and dimensions (attributes). Once defined, these are exposed via REST and GraphQL APIs, enabling any downstream tool — Metabase, Grafana, Superset, a custom React dashboard — to query governed, consistent data.
Cube also handles pre-aggregations (caching query results in the data warehouse for fast reads), multi-tenancy (row-level security and data isolation across customers), and access controls. For engineering teams building analytics products or BI infrastructure, Cube solves the problem of having metric logic scattered across dozens of SQL files and BI tool configurations.
The open-source version (cube.js) is fully featured and free to self-host. Cube Cloud is the managed offering with production deployment tooling, monitoring, and enterprise features. Cloud pricing is consumption-based, measured in Cube Consumption Units.
What Was Mode Analytics?
Mode Analytics was a San Francisco-based SQL analytics platform founded in 2013. It was built for data analysts who wanted to write SQL, visualize results, and share analyses without leaving a single, collaborative environment. Mode's core product was a SQL notebook where analysts could write queries, run Python or R code, build charts, and publish the resulting report to stakeholders — all in one workflow.
Mode differentiated itself from pure BI tools like Tableau or Looker by embracing code-first analysis. Business analysts at Mode were expected to know SQL. Mode's community report library — a public collection of data analyses across industries — was a genuinely valuable resource that drove organic growth.
In June 2023, ThoughtSpot announced the acquisition of Mode Analytics for $200 million. The transaction closed in July 2023. ThoughtSpot's rationale was clear: Mode's code-first analyst workflow complemented ThoughtSpot's natural language and AI-driven exploration, creating a platform that could serve both technical and non-technical users. Mode's capabilities were integrated into ThoughtSpot Analyst Studio, which became generally available to ThoughtSpot Cloud customers in early 2025. Mode as a standalone product is no longer available to new customers.
Side-by-Side Comparison
| Category | Cube (2026) | ThoughtSpot Analyst Studio (Mode Successor) |
|---|---|---|
| Product Status | ✓ Active | ~ Mode discontinued; Analyst Studio active |
| Layer in Stack | Semantic layer / API | Analytics interface / notebook |
| SQL Notebooks | ✗ Not native | ✓ Core capability |
| Python / R Support | ✗ | ✓ Inherited from Mode |
| Metric Governance | ✓ Core strength | ~ Via ThoughtSpot model |
| Pre-aggregations / Caching | ✓ | ✗ |
| Multi-tenancy / Row Security | ✓ | ~ Via ThoughtSpot RLS |
| Open Source Option | ✓ cube.js | ✗ |
| AI Query Assistance | ~ Emerging | ✓ ThoughtSpot AI core |
| Dashboard Building | ✗ Not native | ✓ |
Pricing Comparison
Self-hosted cube.js. Full semantic layer capability. No cost beyond infrastructure. Requires DevOps capacity to deploy and maintain.
Consumption-based. Starter at $0.15/CCU; Premium at $0.30/CCU. Typical cloud deployments run $3,000–$8,000/month. Enterprise requires custom negotiation.
Mode's pricing is no longer relevant — the product is discontinued. Historical plans ranged from approximately $999 to $3,000+ per month for team plans.
ThoughtSpot pricing starts at approximately $1,250/month for small teams. Enterprise packages with full Analyst Studio access are quoted individually. Annual commitment required.
Core Capability Comparison
Semantic Layer and Metric Governance
Cube is the stronger choice here, and it is not a close comparison. The entire purpose of Cube is to create a single, governed definition of every metric your organization tracks. Whether that metric is queried by a Metabase dashboard, a React front-end, or a Grafana monitoring panel, it returns the same number calculated the same way. ThoughtSpot has its own data model and governance capabilities, but they are oriented toward BI exploration rather than serving as a centralized metrics API for engineering teams.
SQL-First Analyst Workflow
ThoughtSpot Analyst Studio wins decisively here. This is what Mode was built for and what ThoughtSpot acquired it to deliver. Analysts can write SQL, run Python cells, iterate on results, build visualizations, and publish polished reports — all in a single notebook environment. Cube has no equivalent analyst-facing interface.
Pre-aggregation and Query Performance
Cube's pre-aggregation system is a meaningful technical differentiator. Cube can materialize frequently-queried results into the data warehouse itself (Snowflake, BigQuery, Redshift), so dashboards load in milliseconds even on very large datasets. This is infrastructure-level performance optimization that BI tools and analytics platforms like ThoughtSpot handle internally but expose less granular control over.
Embedded Analytics
Cube is commonly used as the backend for embedded analytics in SaaS products. When a software company wants to embed charts and metrics into their product for their customers, Cube handles multi-tenancy, access control, and API delivery. ThoughtSpot also has an embedded analytics product (ThoughtSpot Everywhere), but Cube is more flexible as a pure API infrastructure layer.
Integration Ecosystem
Cube integrates with all major data warehouses: BigQuery, Snowflake, Redshift, Databricks, Athena, PostgreSQL, and others. On the downstream side, Cube connects to Metabase, Grafana, Apache Superset, Retool, Tableau, Power BI, and effectively any tool that can query a REST or GraphQL API. This flexibility is a core part of Cube's value proposition.
ThoughtSpot Analyst Studio connects to the same set of warehouses and inherits ThoughtSpot's broader integration ecosystem, including Slack alerts, Google Sheets export, and various enterprise SSO providers. Its integration depth is strong but narrower than Cube's API-first approach.
Ease of Use
These tools serve different users, so "ease of use" means something different in each context. For data analysts, ThoughtSpot Analyst Studio has a familiar SQL notebook interface that most practitioners can use productively on day one. The AI query assistance built into ThoughtSpot reduces friction for less SQL-fluent team members.
Cube requires data engineering knowledge to set up and maintain. Defining a well-structured data model in Cube's YAML or JavaScript schema requires understanding of data modeling concepts, SQL, and ideally some experience with semantic layer design. This is not a tool a business analyst picks up on day one. However, once the model is in place, downstream consumers — including non-technical stakeholders using BI tools connected to Cube — benefit without needing to understand the underlying complexity.
Best Use Cases by Stage
Early-Stage Teams
Early-stage teams rarely need either of these tools at the infrastructure level. A single BI tool connected directly to the data warehouse is almost always sufficient. Cube adds value when metric definitions start diverging across tools. ThoughtSpot Analyst Studio is priced above what most early-stage teams should spend on analytics infrastructure.
Growth-Stage Data Teams
This is where Cube often earns its place. When a data team of 2-5 people is maintaining metric definitions in five different places — dbt, Looker, Metabase, a Retool dashboard, and a custom React page — the maintenance cost becomes unsustainable. Cube centralizes those definitions. ThoughtSpot Analyst Studio serves growth-stage analysts who want a collaborative notebook environment without paying for Databricks or Snowpark notebooks.
Enterprise
At enterprise scale, both tools are relevant — but they often coexist rather than compete. A mature data organization might use Cube as the semantic layer feeding multiple downstream BI tools, while analysts use ThoughtSpot Analyst Studio for exploratory work that does not fit neatly into a governed metric definition. The two tools solve different problems.
The Operating Intelligence Alternative (Fairview)
The Layer Above the Semantic Layer
Cube governs how metrics are defined. ThoughtSpot Analyst Studio is where analysts explore those metrics. But neither tool answers the operator's question: what do I do next?
Fairview is the operating intelligence layer above the stack. It is not a BI tool and not a semantic layer — it is the platform that connects data from your BI tools, CRMs, automation workflows, and financial systems into unified margin and revenue visibility. COOs, founders, and operators use Fairview to understand what is making money, what is leaking margin, and what action to take — without building another dashboard or waiting for a data analyst to run a query.
Alternatives to Consider
dbt Semantic Layer
dbt's native semantic layer allows teams already using dbt for data transformation to define metrics in the same codebase. Tightly integrated with the dbt ecosystem and downstream tools like Metabase and Lightdash. A strong alternative to Cube for teams with existing dbt investment.
AtScale
A commercial semantic layer platform with a strong emphasis on enterprise governance and compatibility with Excel and Power BI. More enterprise-focused than Cube, with a different integration priority.
Lightdash
An open-source BI tool that uses dbt as its data model, providing a governed analytics interface on top of existing dbt transformations. Serves a similar role to the Cube + Metabase combination but in a more integrated package.
Apache Superset
Open-source BI platform with SQL editor, dashboard builder, and a growing semantic layer (Superset datasets). A free alternative for teams that want SQL-first exploration without the cost of ThoughtSpot.
Final Verdict
The Bottom Line
Comparing Cube and Mode Analytics is somewhat like comparing PostgreSQL to Tableau — they are different infrastructure components. Cube is data infrastructure for defining and serving governed metrics. Mode was an analyst workspace for SQL-based exploratory analysis, a role now filled by ThoughtSpot Analyst Studio.
If your team needs to centralize metric definitions across multiple downstream tools and enforce consistency — choose Cube. If your data analysts need a collaborative SQL notebook with visualization and reporting — evaluate ThoughtSpot Analyst Studio, or consider open-source alternatives like Apache Superset or Lightdash. And if your leadership team needs operating intelligence that turns all of that data into clear decisions about margin and revenue — Fairview belongs above both.