Overview
A headless-core React data table where every feature works on the client and on the server.
Introduction
react-tablekit renders tabular data in React. The engine is framework-agnostic and renders nothing on its own; the components on top of it are a default you can replace part by part.
Its organising idea is that where the work happens should not change your code. Pagination, global search, column filters, sorting, grouping, expansion, selection and faceting each run either in the browser or on your server, chosen per feature. Starting client-side and moving to the server later is a prop, not a rewrite.
What it is not: a spreadsheet, a chart library, or a data-fetching library. It has no opinion
about how rows reach it: fetch, React Query, a router loader or an array in memory all work.
Why react-tablekit
- Client and server are the same code. Each feature picks its side; the table handles the debouncing, aborting, race conditions and out-of-range pages that make server mode tedious.
- Nothing is sealed. Every part is a slot, every interaction is middleware you can wrap, and every visual value is a CSS variable.
- Headless underneath.
useDataTableis the whole engine with no markup;createTableruns without React at all. - Accessible by default. Correct table semantics, the right ARIA role for the features you enable, the data-grid keyboard pattern and polite live announcements.
- Zero runtime dependencies. Only React as a peer. ESM and CJS with types for both, tree shakeable and server-render safe.
Start now
- Installation: install the package and the stylesheet.
- Quickstart: a complete first table, nothing elided.
- All features: every capability, each with a live demo.
- Client and server: the defining feature, in one page.
- Customization: the nine levels, from a token to the core.
- API reference: generated from the source, so it cannot drift.
Licence
MIT, with no paid tier and no feature gating. Everything documented on this site is in the package you install.