Hex is an AI analytics platform that brings notebooks, conversational self-serve, and data apps into one governed workspace. It is built for data teams that want to move fast without losing control, and for business users who need answers without writing SQL or Python.
Hex replaces the usual tangle of separate notebooks, BI dashboards, and ad-hoc Slack requests with one connected environment. Data practitioners can run deep analysis in agentic notebooks that support SQL and Python cells, while business users can ask questions in plain language directly in Hex, in Slack, or through MCP. When a question needs deeper investigation, a data practitioner can jump into the same workflow without switching tools. This shared space is designed to reduce tool sprawl and keep everyone working from the same data.
The Context Engine and Trust
The platform's context engine combines database descriptions, semantic models, business rules, and analytical logic into a curated layer that powers every AI answer. Data teams can observe what is being asked, catch inconsistencies, and improve context over time. Built-in evals let teams test changes before production and measure their impact on agent accuracy. The goal is to make AI-generated insights more trustworthy and repeatable by grounding them in the organization's actual data and definitions.
Hex notebooks are designed to be living documents. Analysts can show their logic step by step, documenting metric definitions, data sources, hypotheses, and decisions. Results can be turned into interactive data apps with visual exploration and drill-down, then shared with stakeholders. Real-time collaboration means teams can explore, iterate, and decide in the same place. Scheduled runs and alerts help keep key metrics visible, and version history provides an audit trail.
Hex is built with enterprise-grade security and governance in mind. It offers OIDC SSO, audit logs, OAuth database connections, and options like Bring Your Own Key and single-tenant deployment. Data permissions can restrict editing and viewing, and organizations can manage user groups and usage reports. For teams in regulated industries, HIPAA compliance is available as an add-on. These controls are meant to let organizations adopt AI and data work without losing visibility into how data is being used.
Hex connects to data warehouses and supports orchestration with Airflow, Dagster, and Prefect. Projects can be exported to GitHub or GitLab for version control and audit trails. A public API allows teams to build custom integrations, and marketplace offerings include AWS Marketplace and Snowflake Marketplace. This extensibility helps Hex fit into existing data stacks rather than requiring a wholesale replacement.
Hex offers a free Community plan and paid Professional, Team, and Enterprise plans. A 14-day free trial of the Team plan is available without a payment card. Medium compute is included on all paid plans, while larger compute profiles are billed per minute, with rates ranging from $0.32 per hour for Large up to $4.06 per hour for A10G GPU. AI features consume credits, and paid plans include credit grants that vary by tier. Students, educators, and qualified non-profits may access discounted or free plans. Exact current prices and credit amounts should be verified on the official pricing page.
Hex is most useful for organizations that want to combine AI-assisted analysis with the rigor of code-based workflows. Product analytics teams can explore user behavior and run experiments, while business users can self-serve common questions. The platform's emphasis on context and evals addresses a common concern with AI analytics: inconsistent or untrustworthy answers. By keeping analysis, apps, and governance in one place, Hex aims to reduce the friction between asking a question and acting on the answer.
Hex AI is the AI layer inside Hex, the analytics workspace built by Hex Technologies — not a chatbot add-on and not one of the several unrelated products that share the Hex name. Hex is a browser-based workspace where data teams explore data, build analyses and publish them as interactive data apps, and its AI features work directly on a company's own warehouse data rather than on pasted samples or public information.
The centrepiece is the Notebook Agent. Hex notebooks mix SQL, Python and no-code cells in a single document, and the agent can write, edit and run those cells for you: describe the analysis you want, and it drafts the queries, builds the charts and explains what it found, pausing for approval before it deletes cells or publishes anything. Because the agent works inside the notebook, its code and outputs stay visible and editable — an analyst can inspect every step instead of trusting a black-box answer.
For everyone outside the data team, Hex offers Threads: ask a question in plain language — in Hex itself or from Slack — and get an answer drawn from the same governed data. A shared Context Engine sits underneath both experiences, combining warehouse schemas, dbt documentation and semantic models, plus the definitions your data team endorses in Context Studio, so the agent uses your company's metric definitions rather than guessing. Hex connects natively to Snowflake, BigQuery, Databricks, Redshift, Postgres and other warehouses and databases, queries data where it lives, and is SOC 2 Type II certified.
That combination is why Hex AI is aimed at data teams who want to open up self-serve analytics without losing control: analysts keep a real notebook and real code, business users get fast answers they can trust, and the data team keeps one place to govern definitions, access and quality.
What is Hex AI?
Hex AI is the set of AI features inside Hex, the analytics workspace from Hex Technologies. It includes the Notebook Agent, which writes and runs SQL and Python inside Hex notebooks, and Threads, which lets anyone ask questions about company data in plain language. Answers are grounded in your warehouse data and your team's governed metric definitions.
Is Hex a BI tool?
Hex covers much of the ground of a BI tool — dashboards, data apps and self-serve questions — but it is built around a code-friendly notebook rather than drag-and-drop reporting alone. Data teams typically adopt it to replace a patchwork of notebooks, one-off queries and a separate BI layer with a single workspace.
Which data warehouses does Hex connect to?
Hex connects natively to Snowflake, BigQuery, Databricks, Redshift and Postgres, along with other warehouses and databases such as ClickHouse, Athena and SQL Server. It also syncs documentation and semantic models from dbt, so existing definitions carry over instead of being rebuilt.
How much does Hex cost?
Hex has a free Community plan for getting started. Paid plans are priced per editor: Professional is $36 per editor per month and Team is $75 per editor per month, with custom Enterprise pricing on top. Viewer access and AI usage credits vary by plan, so check the current pricing page before budgeting.