Documents and notes decision

GitBook Premium

A competent developer can replace the core public-site, search, and hosting workflow in ~30 hours, but GitBook’s embedded AI assistant, analytics/insights, and enterprise compliance/migration services are nontrivial to replicate, so keeping the paid product makes sense for those features.

Visit website
You pay

$65/mo

$780/yr

Per seat. Read off the official pricing page.

You’d pay instead

$50one-off30 h to build

$40/mo5 h/mo upkeep

On cash alone, building overtakes the subscription at 1 seat.

Open-source builds that already do this

Every project below is open source and already does this job today. Fork one, self-host it, or take the parts you need - the build prompt further down assumes an empty file, and this is the shortcut past that. Licences differ; check the one on each card before you ship. All GitBook Premium alternatives, with the arithmetic →

What a replacement has to do

  • Publish Markdown docs to a branded public site with full-text search and basic analytics.

What it still won’t have

  • GitBook AI Assistant (embedded advanced chatbot)
  • AI Insights and GitBook Agent features
  • GitBook MCP server & built-in AI optimizations (MCP/llms.txt automation)
  • Enterprise compliance & white‑glove migration
  • Hosted editor/collaboration with built-in merge workflow and team UX

What remains hard

  • Compliance and regulationSOC 2, ISO 27001, SAML SSO, access controls, white-glove migration.
Read the build prompt

First-year cost

Keep paying

Paying is—cheaper in year one.

On cash alone, building overtakes the subscription at 1 seat.

Paid seatsseats

Money you would actually spend

Keep paying
—

Subscription price × seats × 12

Build it
—

AI build —APIs + hosting —

Time you would spend

—

—

What you would spend

What we assumed

The verdict above measures whether you could build it. This one is only about money.

Runnable build prompt

Not run yet
Build a minimal hosted docs platform using Docusaurus (React) + Node.js backend, deploy on Vercel, and use Algolia (or Typesense) for search. Include: Markdown import from a Git repo, static site generation, CI/CD deploy to Vercel, search indexing and client search UI, custom domain setup, simple Git-based collaborator workflow, and basic analytics (Plausible or GA). Out of scope: training/custom LLMs, an embedded AI chat assistant, enterprise SSO onboarding, and white‑glove migration. Provide error handling for imports/CI failures, automated tests for build and search indexing, and deployment/run instructions.
How we checked5 sources · 2/3 runs agreed · evidence score 60

How the score was reached

  • Partly verdict base52
  • An open-source build was found+5
  • 5 cited sources+3
  • Price verified on pricing page+3
  • Hard moats found in the evidence-3
  • Evidence score60

The base comes from the verdict. Everything under it is a check that either happened or did not, and each one is a fact frozen in this record rather than a judgement made at render time - so the same evidence always produces the same number.

How scoring works →

Cited sources · 5

Every page the run actually retrieved.

Integrity checks

What held up, and what did not.

✓ Price read off the page! 2 of 3 runs agreed; the verdict is the majority✓ Citations limited to fetched pages! 1 moat quoted from the page