Observable is a browser-based notebook platform for people who work with data and code. It combines JavaScript, SQL, and prose in a single document, so analysts, developers, journalists, and educators can explore a dataset, build a visualization, and publish an explanation without leaving the browser.
Observable notebooks are made of cells that run in dependency order. Change a value in one cell, and every cell that depends on it updates automatically. That reactive model makes notebooks well suited to iterative work: you can adjust a parameter, swap a dataset, or rewrite a query and see the downstream charts and tables refresh in place. The format is lightweight enough to sketch an idea quickly, but flexible enough to support a full analysis.
Observable has deep ties to the JavaScript visualization ecosystem. D3 and Plot are both part of the same lineage, and notebooks can pull in those libraries alongside other JavaScript packages. The platform also supports database connections, so a notebook can query live data with SQL and feed the results directly into a chart. File attachments let you load datasets into a notebook, and the reactive model means a new query result propagates through the document without manual reruns.
Beyond analysis, Observable is designed for communication. Notebooks can include sliders, inputs, and other interactive controls, which makes it possible to build explorable explanations where readers adjust variables and watch the output change. The same building blocks support custom apps and data-driven workflows. Public notebooks can be embedded in external sites, and the Free plan includes a watermark on embeds.
Observable offers team workspaces that centralize data sources, component libraries, and other shared assets. That structure supports coordination across a team without duplicating setup work in every notebook. The platform also includes an agent feature with daily token allowances that vary by plan, positioning notebooks as a place where autonomous AI agents can assist with data questions alongside human collaborators.
Notebooks use a human-readable, plain-text file format, and Observable provides free open-source tooling for generating static sites. That combination matters for teams that want to keep their work portable or publish notebooks outside the hosted platform. The open-source projects include Notebook Kit, Plot, and D3, which are maintained separately from the commercial product.
Observable has a Free plan for individuals, a Pro plan at $22 per month or $25 billed monthly, a Team plan at $40 per editor per month or $45 billed monthly, and custom Enterprise pricing. Compute is capped at 50 minutes per month on Pro and 50 minutes per editor per month on Team. Private viewers on the Team plan cost $10 per viewer per month. Agent token allowances differ by tier, and Enterprise adds SSO, audit logs, and dedicated support. Check the official pricing page for current details.
Observable works best for people who are comfortable writing code and want a fast path from data to a published, interactive result. It is less suited to teams that need a spreadsheet-style interface or a fully managed BI dashboard. For analysts, developers, and educators who treat code as the primary tool for thinking about data, the reactive notebook model and the surrounding visualization ecosystem are the main reasons to consider it.