Hex iconHex

AI analytics workspace with a Notebook Agent, conversational Threads, semantic models and SQL plus Python notebooks for data teams

$0 Community Plan
4.5/5 G2 Rating
$70M Series C (May 2025)

Overview

Hex calls itself "The Analytics platform for humans and agents". Data teams work in notebooks that mix SQL, Python and charts, and publish them as apps; the Notebook Agent writes queries, code and charts, and Threads lets business users ask questions in plain language and get answers built on the team's semantic models. Context Studio, Evals and a Hex MCP server help data teams control what the agent sees and check its answers. Recent releases include a governed mode that limits the agent to approved tables and models (October 2026) and a Hex 4.0 public beta (November 2026). Hex Technologies was founded in 2019 in San Francisco and raised a $70M Series C led by Avra in May 2025; it says more than 2,000 companies use it. For code-only notebooks see Deepnote and Jupyter.

The Verdict

Who Should Use Hex?

Best For

  • Data teams fluent in SQL and Python that want an AI agent in their notebooks
  • Teams that want to publish analyses as apps for the rest of the company
  • Companies that want governed self-serve questions on a semantic layer

Not Ideal For

  • Business users with no data team behind them
  • Teams that need pixel-perfect dashboards and heavy chart customization

What's Great

  • Ease of use and data integrations lead G2 praise
  • SQL and Python work side by side in one notebook
  • Free Community plan and a 14-day Team trial without a card

Watch Out For

  • Slow with many tables and heavy compute, per G2 reviewers
  • Large projects get hard to organize; limited chart customization
  • A competitor's review rates it weaker for broad self-serve use

Pricing

View all features & details

AI Features

  • Notebook Agent
  • Threads (conversational analytics)
  • Context Studio and semantic models
  • Evals for agent answers
  • Hex MCP server, Admin API and CLI

Platform

  • SQL and Python notebooks
  • Published data apps
  • Explore for point-and-click analysis
  • GPU and large compute add-ons
  • Free Professional plan for students and educators

How It Compares

Feature Hex Julius ThoughtSpot
Main user Data teams Individuals and small teams Business users, governed by data teams
Free plan Yes (Community) Yes, limited No (free trial; free embedded developer plan)
Paid entry price $36/editor/mo $45/mo (Pro) $25/user/mo (Essentials, annual)
Code access SQL and Python notebooks Generated Python, R, SQL No code needed

Sources: Hex · Hex about · Hex changelog · Julius review (Feb 2026) · ThoughtSpot pricing

Is this your tool? Add a badge