A Notebook War Playing Out in Plain Sight
Hex Technologies has spent the last two years building something that looks, on the surface, like a better Jupyter notebook. Underneath, it is a direct challenge to the collaborative data science workflow that Deepnote spent years making its core identity. The two tools overlap enough that teams choosing one are almost always choosing against the other – and right now, that choice is trending in one direction.
Deepnote arrived early to the cloud-native notebook space, carving out a loyal following among data science teams that wanted real-time collaboration without the friction of local environments. Hex is now targeting that exact base, not by undercutting on price, but by offering something Deepnote has struggled to deliver cleanly: a path from exploratory analysis all the way to a published, interactive data app – without leaving the notebook.
What Hex Is Actually Selling
Hex positions itself as a “data workspace,” which is a deliberately broad term. In practice, it means a collaborative SQL and Python notebook that can publish finished work as a shareable app or dashboard. That last step – the publishing layer – is where Hex pulls away from the notebook-only model that Deepnote built its reputation on.
The workflow Hex enables is specific: an analyst writes SQL to pull data, switches to Python for modeling or cleaning, and then surfaces results through an interactive interface that non-technical stakeholders can actually use. The whole chain happens in one tool. Teams that previously stitched together a notebook, a BI layer, and a sharing solution are collapsing that stack.
This is not a marginal improvement in convenience. Data teams are under real pressure to show their work to business partners who will not open a Jupyter notebook and who find most BI dashboards too rigid to answer follow-up questions. Hex’s app publishing layer sits exactly in that gap. It is not as powerful as a full BI tool, but for ad-hoc analysis with interactive controls, it removes a step that many teams found genuinely painful.
Where Deepnote Still Holds Ground
Deepnote is not standing still. Its collaboration features – real-time co-editing, inline comments, version history – remain strong, and its integrations with data warehouses like Snowflake and BigQuery are mature. For teams that are deeply notebook-centric and do not need a publishing layer, Deepnote’s experience is cleaner and less opinionated about how work should flow.
The problem for Deepnote is that “notebook-centric” describes fewer teams every year. As data functions grow inside companies, the pressure to share findings beyond the data team increases. Deepnote does offer sharing and app-like views, but the feature has not become a core part of its identity the way the publishing layer has for Hex. That positioning gap is where Hex is doing its most effective recruiting.
The Team Composition Factor
One pattern worth tracking: Hex adoption tends to accelerate when a data team sits close to a product or growth function. Those are the environments where analysts are regularly expected to hand off interactive reports rather than static outputs, and where the audience for data work includes people who will ask questions a fixed chart cannot answer. Hex’s app layer answers that directly.
Deepnote’s strongest retention, by contrast, tends to be in research-oriented teams or academic settings where the notebook format is the output, not a step toward something else. That is a real and defensible niche, but it is smaller than the enterprise analytics market Hex is pursuing. The center of gravity in data tooling has been moving toward “insights for everyone” for several years, and Hex is better aligned with that direction right now.
The SQL-first angle also matters more than it initially appears. A large share of working data analysts are more comfortable in SQL than Python, and tools that make SQL a first-class citizen rather than an afterthought tend to see faster team-wide adoption. Hex built SQL cells as a genuine part of the notebook experience rather than bolting them on. Deepnote supports SQL, but the tool’s personality is still Python-native in a way that can feel like friction for SQL-heavy teams.
Pricing structure adds another layer to the competitive picture. Hex’s free tier is generous enough that individual analysts can build meaningful workflows before an organization ever pays, which mirrors the bottom-up growth strategy that has worked well for tools like Notion – where power users drive adoption before IT gets involved. Once a handful of analysts inside a company are publishing Hex apps to business partners, the case for a team upgrade writes itself. Deepnote uses a similar individual-first model, but Hex’s app publishing creates a more visible artifact that spreads organically across an organization.
The real question is not whether Hex is winning users from Deepnote – it clearly is – but whether it can hold them as the data tooling market continues to consolidate. Databricks, dbt Labs, and a growing number of enterprise platforms are all pushing into the collaborative notebook space from different angles. Hex’s window as an independent, focused alternative may be narrower than its current momentum suggests.
