Jupyter Book

Overview

Jupyter Book turns technical writing and notebooks into computational publications that can be built for the web or other formats. It is best suited to researchers, educators, and data scientists who need explanations and executable work in one place. Its flexibility is a strong fit for that purpose, but its two-version transition calls for a deliberate choice of workflow.

Key features

Authors can combine MyST Markdown and Jupyter notebooks with citations, cross-references, figures, tables, equations, and code blocks. This makes Jupyter Book useful when readers need to follow both an explanation and the computation behind it; it is less compelling for documentation that is only prose. Books can be built as websites, PDFs, or other formats and published to the web.

Code can run during a build to generate figures, tables, or values for publication, but execution is off by default. That avoids running code as an automatic part of every build, while requiring authors who want computed results to enable execution. Readers can also encounter interactive widgets, run code in a browser through a remote kernel, or launch Binder and JupyterHub environments.

Plugins can add directives, roles, and content transformations, and can be distributed as files, npm packages, or release artifacts. Installation is available through PyPI, conda-forge, and npm, with methods including pip, mamba or conda, pipx, uv, and pixi. A local preview server updates as content changes, which helps authors review a book while writing. Deployment options include GitHub Pages, Read the Docs, GitLab Pages, Netlify, Curvenote, and custom web servers.

Pricing

Jupyter Book is open-source software and costs nothing. There is no paid plan or free trial to weigh against that price. The trade-off is that users work with an installable project rather than a priced hosted service; custom domains are supported, while documentation versioning is not.

Platforms

Jupyter Book supports Linux, macOS, and Windows, as well as web and self-hosted use. Its installation routes and hosting choices suit teams that want to work locally and publish through infrastructure they choose.

Who it's for

Researchers, educators, and data scientists are the clearest audience: the format brings narrative, code, and computed results together for reusable computational publications. It can also suit technical teams whose documentation benefits from notebooks, equations, or interactive learning. Teams needing conventional API documentation or a stable Sphinx-extension workflow should weigh the version 2 gaps carefully.

Pros and cons

• Pros: One project supports notebooks, technical prose, citations, equations, and executable content, so authors can publish an explanation alongside its computational basis.
• Pros: Multiple output formats, interactive options, plugins, and varied hosting routes give authors room to shape both the publication and its deployment.
• Cons: Jupyter Book 2 is a rewrite on the MyST Document Engine and the project says it still needs many improvements; execution parity, editable Thebe cells, Sphinx autodoc-style API docs, and exporting notebook outputs as MyST are not yet available.
• Cons: Custom Sphinx extensions do not work with version 2, and books that directly use Sphinx's conf.py cannot be automatically upgraded. Meanwhile, version 1 is mostly in maintenance mode, without planned active development for its tools and with limited review commitments for community contributions.
• Cons: No security certifications or compliance claims are stated, which may make it a poor fit for buyers who require such assurances.

Alternatives

Sphinx is a reasonable choice when an existing documentation workflow depends on Sphinx extensions: it is free, open-source software that can be installed locally or self-hosted.

Fern may fit teams seeking a hosted documentation product with a free allowance; its free plan covers 10 team members, 1,000 pages, 250 AI credits per month, 2,000 page generations per month, and one site.

Mintlify offers a free Starter plan with five editor seats, a custom domain, web editor, authentication, MCP server, and API playground, making it an option for teams that want those included product features.

VitePress is a free MIT-licensed static site generator for readers comfortable with its Node.js 22-or-higher requirement. DocFX is another free static site generator, requiring .NET SDK 8.0 or higher.

Dokiel is free and open source, distributed as a Scenari publishing chain for desktop or SCENARIchain installation. GitBook has a free plan for individuals, with one free user, unlimited page views, and 10 GitBook Agent messages per week. Dr.Explain has a Windows free plan limited to 20 topics.

Verdict

Choose Jupyter Book if you need open-source computational publications that unite notebook content, explanatory writing, and flexible publishing. Its strongest reason to choose is that breadth without a software charge; its main reason to look elsewhere is the transition between a version 1 stack in maintenance mode and a version 2 rewrite with meaningful compatibility and feature gaps.