Build and deploy scalable products on Cloudflare with ease.
An opinionated, batteries-included starter kit for quickly building and deploying SaaS products on Cloudflare. This is a Next.js project bootstrapped with c3
.
This is the same stack used to build Supermemory.ai which is open source at git.new/memory
Supermemory now has 20k+ users and it runs on $5/month. safe to say, it's very effective.
- Next.js for frontend
- TailwindCSS for styling
- Drizzle ORM for database access
- NextAuth for authentication
- Cloudflare D1 for serverless databases
- Cloudflare Pages for hosting
- ShadcnUI as the component library
-
Make sure that you have Wrangler installed. And also that you have logged in with
wrangler login
(You'll need a Cloudflare account) -
Clone the repository and install dependencies:
git clone https://github.com/Dhravya/cloudflare-saas-stack cd cloudflare-saas-stack npm i -g bun bun install bun run setup
-
Run the development server:
bun run dev
Open http://localhost:3000 with your browser to see the result.
Besides the dev
script, c3
has added extra scripts for Cloudflare Pages integration:
pages:build
: Build the application for Pages using@cloudflare/next-on-pages
CLIpreview
: Locally preview your Pages application using Wrangler CLIdeploy
: Deploy your Pages application using Wrangler CLIcf-typegen
: Generate typescript types for Cloudflare env.
Note: While the
dev
script is optimal for local development, you should preview your Pages application periodically to ensure it works properly in the Pages environment.
Cloudflare Bindings allow you to interact with Cloudflare Platform resources. You can use bindings during development, local preview, and in the deployed application.
For detailed instructions on setting up bindings, refer to the Cloudflare documentation.
Quick explaination of D1 set up:
- D1 is a serverless database that follows SQLite convention.
- Within Cloudflare pages and workers, you can directly query d1 with client api exposed by bindings (eg.
env.BINDING
) - You can also query d1 via rest api
- Locally, wrangler auto generates sqlite files at
.wrangler/state/v3/d1
afterbun run dev
. - Local dev environment (
bun run dev
) interact with local d1 session, which is based on some SQlite files located at.wrangler/state/v3/d1
. - In dev mode (
bun run db:<migrate or studio>:dev
), Drizzle-kit (migrate and studio) directly modifies these files as regular SQlite db. Whilebun run db:<migrate or studio>:prod
use d1-http driver to interact with remote d1 via rest api. Therefore we need to set env var at.env.example
To generate migrations files:
bun run db:generate
To apply database migrations:
- For development:
bun run db:migrate:dev
- For production:
bun run db:migrate:prd
To inspect database:
- For local database
bun run db:studio:dev
- For remote database
bun run db:studio:prod
Don't forget to add the CORS policy to the R2 bucket. The CORS policy should look like this:
[
{
"AllowedOrigins": [
"http://localhost:3000",
"https://your-domain.com"
],
"AllowedMethods": [
"GET",
"PUT"
],
"AllowedHeaders": [
"Content-Type"
],
"ExposeHeaders": [
"ETag"
]
}
]
You can now even set up object upload.
If you prefer manual setup:
- Create a Cloudflare account and install Wrangler CLI.
- Create a D1 database:
bunx wrangler d1 create ${dbName}
- Create a
.dev.vars
file in the project root with your Google OAuth credentials and NextAuth secret.AUTH_SECRET
, generate by commandopenssl rand -base64 32
orbunx auth secret
AUTH_GOOGLE_ID
andAUTH_GOOGLE_SECRET
for google oauth.- First create OAuth consent screen. Tips: no wait time if you skip logo upload.
- Create credential. Put
https://your-domain
andhttp://localhost:3000
at "Authorized JavaScript origins". Puthttps://your-domain/api/auth/callback/google
andhttp://localhost:3000/api/auth/callback/google
at "Authorized redirect URIs".
- Generate db migration files:
bun run db:generate
- Run local migration:
bunx wrangler d1 execute ${dbName} --local --file=migrations/0000_setup.sql
or using drizzlebun run db:migrate:dev
- Run remote migration:
bunx wrangler d1 execute ${dbName} --remote --file=migrations/0000_setup.sql
or using drizzlebun run db:migrate:prod
- Start development server:
bun run dev
- Deploy:
bun run deploy
- Fully scalable and composable
- No environment variables needed (use
env.DB
,env.KV
,env.Queue
,env.AI
, etc.) - Powerful tools like Wrangler for database management and migrations
- Cost-effective scaling (e.g., $5/month for multiple high-traffic projects)
Just change your Cloudflare account ID in the project settings, and you're good to go!