Brandfine SDK v0.1 — public preview

Build content-driven products on Brandfine

A multi-tenant headless CMS with localized content, structured navigations, and instant cache invalidation via webhooks. Everything you need to wire up a marketing site, a SaaS dashboard, or a content portal.

Get started in 60 seconds →Read the docs

What is Brandfine?

A headless CMS that treats your frontend as a first-class consumer. Structured content, typed APIs, and an SDK that handles the boring parts (auth, caching, locale resolution).

Multi-tenant by design

Every team has isolated workspaces. API keys, content, and webhooks are scoped — no cross-tenant leakage.

Localized content

First-class locales. Translation groups link sibling posts; the SDK resolves the right one per request.

Webhook-driven SWR

Server-side caches refresh the moment editors publish. Stale-while-revalidate keeps requests fast.

Pick your path

Same data, two integration shapes. Use the SDK if you're on Node-flavored TypeScript; drop to REST anywhere else.

TypeScript SDK

Recommended

Typed client + server-side caches + locale resolvers + a one-line webhook handler. Works on Astro, Next, Remix, Bun, Workers.

npm i @brandfine/client

Read the SDK guide→

REST API

Plain HTTP under /external/*. Bring your own client, your own caching layer, your own language. Bearer-token auth, JSON in and out.

curl -H 'Authorization: Bearer …' …

Read the API reference→

60-second start

From zero to fetching content

Three steps. Create a workspace API key in the dashboard, install the SDK, fetch your first post. The SDK ships with stale-while-revalidate caches, so production reads are served from memory.

Full getting-started guide →

src/lib/bf.ts

import { createBrandfineClient } from '@brandfine/client'

export const bf = createBrandfineClient({
  baseUrl: process.env.BRANDFINE_API_URL!,
  apiKey: process.env.BRANDFINE_API_KEY!,
})

const posts = await bf.posts.list({
  type: 'blog',
  locale: 'en',
})

For AI assistants

Using ChatGPT, Claude, or Cursor?

We ship llms.txt files so AI tools can ingest these docs without scraping styled HTML. Drop the URL below into your prompt — the model picks up the entire reference in one shot.

llms.txtIndex — categorized links to every page

→

llms-full.txtFull docs as one markdown file (≈1.3k lines)

→