Hex

What is Hex?

Hex is a browser-based AI analytics platform for working with data from warehouses, databases, files, and other connected sources. Its central workspace combines SQL, Python, charts, markdown, spreadsheet-style calculations, no-code exploration, and AI assistance in the same project.

The product is designed to bridge the gap between technical analysis and self-service analytics. An analyst or data scientist can inspect and transform data in a notebook, while a business user can interact with the resulting application or ask questions about governed data without editing the underlying code.

How Hex is used in practice

A typical workflow starts by connecting Hex to a warehouse or database, or by importing a supported file or notebook. The user then creates a project, explores the data with SQL, Python, visualizations, or no-code cells, and uses AI where it can accelerate analysis. The completed project can be shared as an interactive app, embedded in another product, or scheduled for recurring runs.

This makes Hex more than a chatbot layered over a database. Its AI operates within an inspectable analytics project, where generated SQL, Python, charts, and transformations can be reviewed and modified. That is useful when the result needs to be reproducible rather than merely conversational.

Core analytics and AI capabilities

Notebooks for technical analysis

Hex notebooks support SQL and Python alongside markdown, charts, pivot tables, spreadsheet-style calculations, and no-code cells. This allows a project to move from data preparation to analysis and explanation without switching between several specialized tools.

The Notebook Agent can generate, edit, explain, and debug SQL, Python, markdown, chart, and pivot cells using project context. It is therefore most useful as an assistant inside an existing analytical workflow, not as a replacement for understanding the data or checking the resulting logic.

Natural-language data exploration

Hex provides conversational features for asking questions about connected data and published applications. Threads and Chat with App can return analyses, charts, tables, spreadsheets, and explanations. The usefulness of these answers depends on the quality of the connected data, permissions, and semantic context supplied to the workspace.

Hex also supports semantic models, endorsed data, workspace rules, and Context Studio. These governance features are intended to give AI and self-service users more consistent definitions and more appropriate access to organizational data.

Interactive data apps

A Hex project can be published as an interactive data application. Instead of sending a static report, a team can give stakeholders an interface with filters and other controls for exploring the analysis. Apps can also be shared through public, private, or signed embedding, with optional row-level security for signed embeds.

Recurring analytical workflows

Paid plans support scheduled project runs and alerts, allowing teams to automate recurring analytical work and keep published outputs current. Hex also connects with data platforms and orchestration workflows, including services such as n8n, Airflow, Dagster, and Prefect, according to the supplied product information.

Connections and integrations

Hex connects to common warehouses and databases, including BigQuery, Snowflake, Redshift, PostgreSQL, MySQL, ClickHouse, Trino, Starburst, Athena, and Databricks. It also supports dbt-related metadata and semantic integrations, GitHub and GitLab workflows, Slack, APIs, MCP clients, and embedded analytics.

These connections are important because Hex is not primarily a standalone dataset or dashboard product. Its value depends on bringing relevant organizational data into a governed environment. Teams already using platforms such as Databricks may use Hex as an analysis and sharing layer, subject to the connection and permission configuration of their workspace.

Who is Hex for?

Hex is mainly suited to data analysts, analytics engineers, data scientists, data engineers, business intelligence teams, and organizations that want business users to work with governed data. It is particularly relevant when a team needs both deep technical analysis and an approachable way to distribute results.

• Data teams: Combine SQL, Python, visualizations, and AI assistance in a shared project.
• Analytics engineers: Add semantic context, permissions, and reusable governed workflows around self-service analysis.
• Business stakeholders: Explore published data apps and ask questions without needing to edit notebook code.
• Product and engineering teams: Embed analytical applications into other web experiences through supported embedding options.

Hex is less suitable for someone who wants a simple consumer chatbot, a mobile-first analytics application, or a dashboard that requires little data-source setup. It is also not primarily a general-purpose writing assistant, customer-support tool, or standalone business intelligence product with no need for technical configuration.

Pricing and access

Hex has a free Community plan with limited usage, small compute, data-source connections, notebook cell types, and limited AI credits. The Professional plan is listed at $36 per Editor per month. The Team plan is listed at $75 per Editor per month and includes a 14-day trial. Enterprise pricing is custom.

Paid plans use AI credits, and the amount consumed can vary with task effort. Plans also differ in published-app limits, compute, scheduling, alerts, governance, and administrative features. Additional compute or credits may create separate charges, so the listed editor price is not necessarily the complete cost for intensive analytical workloads.

Privacy, governance, and setup considerations

Hex states that it does not sell customer data and that its model providers do not train on customer data. Its documented default arrangement uses zero data retention for providers, although certain advanced models may require limited retention for safety and security monitoring when enabled by an administrator. Retention and processing can therefore vary according to the model and workspace configuration.

AI requests may include relevant workspace inputs and outputs, project code, run data, prompts, metadata, schema information, and table or column names. Teams should review which data sources are connected and how permissions are configured before enabling AI features on sensitive projects.

Hex documents SOC 2 Type II and HIPAA compliance, GDPR and CCPA compliance, workspace permissions, and data-source controls. Enterprise features include options such as audit logs, OIDC single sign-on, BYOK, and additional governance controls. Availability of these features depends on plan level or add-ons.

Strengths and limitations

Hex’s main strength is the combination of technical and business-facing analytics in one workflow. A project can contain real code and inspectable analysis while still being published as an interactive application. Its AI features are more useful in this context than a generic answer interface because they can work with project structure, connected data, and semantic context.

The trade-off is complexity. Teams generally need to connect data sources, establish permissions, define governance practices, and understand the resulting SQL or Python. AI usage is credit-based, advanced functionality is concentrated in paid and Enterprise plans, and the primary experience is web-based rather than a native mobile or desktop application.

Is Hex a good fit?

Hex is a strong fit for organizations that want governed self-service analytics without separating notebook work from stakeholder-facing applications. It is especially appropriate when analysts need SQL and Python but business users need simpler ways to explore the results.

It is a weaker fit for casual users seeking instant answers with no setup, teams that only need static dashboards, or organizations without accessible and well-managed data sources. The platform can reduce friction between analysis and sharing, but it does not remove the need for sound data modeling, permission design, and human review of AI-generated work.