// the take
Slate is a React-based framework for building rich text editors, where instead of getting a ready-made WYSIWYG you get a schema-less document model and a plugin API to build your own. It's aimed at teams building a Medium- or Notion-style editor with custom blocks, not teams who just want a drop-in toolbar.
The document model is a nested, DOM-like tree rather than a flat list, so nested structures like tables or blockquotes-within-blockquotes are a first-class case, not a hack. Schema-less core plus a real plugin architecture means you're not fighting baked-in assumptions the way you do with some older editors. The monorepo splits cleanly into slate (data model), slate-react (view), slate-dom, and slate-history, so the pieces you don't need don't ride along. Active maintenance is visible in the code itself — there's dedicated handling for Android IME composition and input quirks, which means someone's actually dealing with contenteditable's worst edge cases instead of pretending they don't exist.
It's been in beta for years with no 1.0 on the horizon, and the README is upfront that breaking changes still happen as the architecture gets reworked — budget for upgrade pain. It's contributor-driven with no company behind it, so if you hit a gap in advanced use cases, the expectation is you PR the fix yourself. Being contenteditable-based means you inherit the usual cross-browser quirks (the amount of Android-specific input code is itself a signal of how much tuning this takes). Collaboration and HTML/Markdown serialization are described as 'ready for' rather than built-in — compared to something like TipTap/Prosemirror with more batteries included, you're wiring up more of that yourself.
We read the README, the directory tree and a key source file in November 2026. Nobody pays to be listed. How we judge →