EditorialWe wrote this page from the company's own website. The company did not ask for it and did not pay for it. Every claim links to the page it came from.
Hex is an AI analytics and BI platform where people ask data questions in plain language and build dashboards, with answers grounded in their company's own data context.[1][3][4]
EditorialUpdated 19 Nov 2026Written by Beacon from Hex's website; passed our listing checks (sources, claims, company domain)Not claimed by the companyMade by Hex
What it does
People can ask data questions in plain language through a feature called Threads, without writing SQL. Questions can range from simple lookups to complex analysis.[3][4]
Answers and findings can be turned into dashboards and data apps from a prompt.[1][3][4]
Context Studio is Hex's governance layer for its AI agent. Data teams use it to ground answers in trusted context, audit how answers were produced, and spot gaps in that context.[1][3]
Hex says its agents can migrate Tableau and Looker dashboards into Hex automatically. Hex says the SQL and logic behind each migrated asset can be inspected.[3]
Hex shows token spend for each conversation thread, so teams can see who is spending what on which projects. It also offers eval sweeps to compare price and quality across AI models.[1]
Who it is for
Business users can explore data and build dashboards without code. Technical users can work in SQL or Python for deeper analysis.[3]
Data teams manage the context the agent uses and review what people ask and how the agent answers.[3][4]
Hex has four plans. Community is for hobbyists. Professional is for individuals. Team is for companies that collaborate on data. Enterprise is for organisations that need advanced security, governance and support.[2]
Hex's website names Notion, Figma, Mercor and Neo as companies using the product.[1]
How it works
The Hex Agent draws on workspace rules, guides, semantic models, and table and column descriptions. It gives priority to tables and projects that the data team has endorsed.[4]
Semantic models can be built in Hex or synced from dbt MetricFlow, Snowflake Semantic Views or Cube. A semantic layer is not required, because the agent can also answer directly from warehouse tables.[3][4]
Every agent conversation is backed by a Hex project. The data team can open that project and inspect the SQL, Python and results behind an answer.[4]
On Team and Enterprise plans, people can @mention the Hex Agent in Slack. Through the Hex MCP server, people can also start conversations from Claude, ChatGPT, Cursor, Glean and other MCP clients.[4]
The agent can run on a schedule and send fresh answers to email or Slack. Hex says it runs on Anthropic and OpenAI models, and its homepage also lists model choice that includes open-source models.[1][4]
Pricing
There is a free Community plan, plus Professional, Team and Enterprise plans. The sources do not list prices for the paid plans.[2]
Anyone can try the Team plan free for 14 days without a payment card. If you do nothing at the end of the trial, the account moves to the free Community plan and loses premium features.[2]
Medium compute is included on all paid plans. Team and Enterprise customers can opt in to pay-as-you-go compute, which is billed per minute of usage.[2]
Professional and Team customers can cancel or change their plan at any time from Settings. Enterprise customers handle renewals and cancellations through their account representative.[2]
Before you buy
Asking the Hex Agent questions requires the Explorer role or higher. Viewers can only see conversations that are shared with them.[4]
Per-user warehouse permissions through OAuth are available only on Enterprise. On other plans, queries run on the connection's shared credentials.[4]
The Slack integration is available only on the Team and Enterprise plans.[4]
Admins can limit the agent to endorsed sources with Endorsed Mode. They can also test the agent against set questions using Evals.[4]
Prices for the paid plans are not shown in the sources, so buyers will need to ask Hex directly.[2]
Alternatives
Other analytics & data tools we have pages for. We do not rank them.