TanStack Start

import { AutoconfigDiagram, Details, Render, PackageManagers, Steps, TypeScriptExample, WranglerConfig, } from "~/components";

TanStack Start is a full-stack framework for building web applications with server-side rendering, streaming, server functions, and bundling.

:::tip[Already have a TanStack Start project?] Run wrangler deploy in a project without a Wrangler configuration file and Wrangler will automatically detect TanStack Start, generate the necessary configuration, and deploy your project.

For more information, refer to Automatic project configuration. :::

<AutoconfigDiagram framework="TanStack Start" configs={[ { label: "main", file: "wrangler.jsonc", value: "@tanstack/react-start/server-entry", }, { label: "compatibility_flags", file: "wrangler.jsonc", value: "nodejs_compat", }, { label: "observability", file: "wrangler.jsonc", value: "enabled: true" }, ]} />

Create a new application

Create a TanStack Start application pre-configured for Cloudflare Workers:

Start a local development server to preview your project during development:

Configure an existing application

If you have an existing TanStack Start application, configure it to run on Cloudflare Workers:

:::note

Cloudflare does not support TanStack Start projects that use Rsbuild. If your application uses Rsbuild, migrate it to Vite before following these configuration steps.

:::

1. Install @cloudflare/vite-plugin and wrangler:

2. Add the Cloudflare plugin to your Vite configuration:

   If your Vite configuration includes another deployment adapter, such as nitro(), remove the adapter and its import before adding the Cloudflare Vite plugin.

   import { defineConfig } from "vite";
   import { tanstackStart } from "@tanstack/react-start/plugin/vite";
   import { cloudflare } from "@cloudflare/vite-plugin";
   import react from "@vitejs/plugin-react";
   
   export default defineConfig({
   	plugins: [
   		cloudflare({ viteEnvironment: { name: "ssr" } }),
   		tanstackStart(),
   		react(),
   	],
   });
   

3. Add a wrangler.jsonc configuration file:

   {
   	"$schema": "node_modules/wrangler/config-schema.json",
   	"name": "<YOUR_PROJECT_NAME>",
   	"compatibility_date": "$today",
   	"compatibility_flags": ["nodejs_compat"],
   	"main": "@tanstack/react-start/server-entry",
   	"observability": {
   		"enabled": true,
   	},
   }
   

4. Update the scripts section in package.json:

   {
   	"scripts": {
   		"dev": "vite dev",
   		"build": "vite build",
   		"preview": "vite preview",
   		"deploy": "npm run build && wrangler deploy",
   		"cf-typegen": "wrangler types"
   	}
   }
   

Deploy

Deploy to a *.workers.dev subdomain or a custom domain from your machine or any CI/CD system, including Workers Builds.

<PackageManagers type="run" args={"deploy"} />

:::note

Preview the build locally before deploying:

:::

Custom entrypoints

TanStack Start uses @tanstack/react-start/server-entry as your default entrypoint. Create a custom server entrypoint to add additional Workers handlers such as Queues and Cron Triggers. This is also where you can add additional exports such as Durable Objects and Workflows.

:::note

If you are using TypeScript, generate Workers types before adding a custom entrypoint:

:::

1. Create a custom server entrypoint file:

   import handler from "@tanstack/react-start/server-entry";
   
   // Export Durable Objects as named exports
   export { MyDurableObject } from "./my-durable-object";
   
   export default {
       fetch: handler.fetch,
   
       // Handle Queue messages
       async queue(batch, _env, _ctx) {
           for (const message of batch.messages) {
               console.log("Processing message:", message.body);
               message.ack();
           }
       },
   
       // Handle Cron Triggers
       async scheduled(event, _env, _ctx) {
           console.log("Cron triggered:", event.cron);
       },
   } satisfies ExportedHandler<Env>;
   

2. Update your Wrangler configuration to point to your custom entrypoint:

   {
   	"main": "src/server.ts",
       "upload_source_maps": true,
   }
   

Test scheduled handlers locally

Test your scheduled handler locally using the /cdn-cgi/local/scheduled endpoint:

curl "http://localhost:3000/cdn-cgi/local/scheduled?cron=*+*+*+*+*"

import { WorkflowEntrypoint, WorkflowStep } from "cloudflare:workers";
import type { WorkflowEvent } from "cloudflare:workers";

export class MyWorkflow extends WorkflowEntrypoint<Env> {
	async run(event: WorkflowEvent<{ input: string }>, step: WorkflowStep) {
		const result = await step.do("process data", async () => {
			return `Processed: ${event.payload.input}`;
		});

		await step.sleep("wait", "10 seconds");

		await step.do("finalize", async () => {
			console.log("Workflow complete:", result);
		});
	}
}

{
	"workflows": [
		{
			"name": "my-workflow",
			"binding": "MY_WORKFLOW",
			"class_name": "MyWorkflow",
		},
	],
}

{
	"services": [
		{
			"binding": "AUTH_SERVICE",
			"service": "auth-worker",
		},
	],
}

import { createServerFn } from "@tanstack/react-start";
import { env } from "cloudflare:workers";

const verifyUser = createServerFn()
	.validator((token: string) => token)
	.handler(async ({ data: token }) => {
		const result = await env.AUTH_SERVICE.verify(token);
		return result;
	});

Bindings

Your TanStack Start application can be fully integrated with the Cloudflare Developer Platform, in both local development and in production, by using bindings.

Access bindings by importing the env object in your server-side code:

import { createFileRoute } from "@tanstack/react-router";
import { createServerFn } from "@tanstack/react-start";
import { env } from "cloudflare:workers";

export const Route = createFileRoute("/")({
	loader: () => getData(),
	component: RouteComponent,
});

const getData = createServerFn().handler(() => {
	// Access bindings via env
	// For example: env.MY_KV, env.MY_BUCKET, or env.AI
});

function RouteComponent() {
	// ...
}

Generate TypeScript types for your bindings based on your Wrangler configuration:

Use R2 in a server function

Add an R2 bucket binding to your Wrangler configuration:

{
	"r2_buckets": [
		{
			"binding": "MY_BUCKET",
			"bucket_name": "<YOUR_BUCKET_NAME>",
		},
	],
}

Access the bucket in a server function:

import { createServerFn } from "@tanstack/react-start";
import { env } from "cloudflare:workers";

const uploadFile = createServerFn({ method: "POST" })
	.validator((data: { key: string; content: string }) => data)
	.handler(async ({ data }) => {
		await env.MY_BUCKET.put(data.key, data.content);
		return { success: true };
	});

const getFile = createServerFn()
	.validator((key: string) => key)
	.handler(async ({ data: key }) => {
		const object = await env.MY_BUCKET.get(key);
		return object ? await object.text() : null;
	});

Static prerendering

Prerender your application to static HTML at build time and serve as static assets.

import { defineConfig } from "vite";
import { cloudflare } from "@cloudflare/vite-plugin";
import { tanstackStart } from "@tanstack/react-start/plugin/vite";
import react from "@vitejs/plugin-react";

export default defineConfig({
	plugins: [
		cloudflare({ viteEnvironment: { name: "ssr" } }),
		tanstackStart({
			prerender: {
				enabled: true,
			},
		}),
		react(),
	],
});

For more options, refer to TanStack Start static prerendering.

Prerendering data sources

:::caution

Prerendering runs at build time. It uses your local environment variables, secrets, and bindings storage data.

:::

To prerender with production data, use remote bindings.

In CI environments, environment variables or secrets may not be available during the build. To make them accessible:

• Set CLOUDFLARE_INCLUDE_PROCESS_ENV=true in your CI environment and provide the required values as environment variables.
• If using Workers Builds, update your build settings.