Vercel
Edit this page on GitHubTo deploy to Vercel, use adapter-vercel
.
This adapter will be installed by default when you use adapter-auto
, but adding it to your project allows you to specify Vercel-specific options.
Usage permalink
Install with npm i -D @sveltejs/adapter-vercel
, then add the adapter to your svelte.config.js
:
ts
importadapter from '@sveltejs/adapter-vercel';export default {kit : {adapter :adapter ({// see the 'Deployment configuration' section below})}};
Deployment configuration permalink
To control how your routes are deployed to Vercel as functions, you can specify deployment configuration, either through the option shown above or with export const config
inside +server.js.ts
, +page(.server).js.ts
and +layout(.server).js.ts
files.
For example you could deploy some parts of your app as Edge Functions...
ts
export constconfig = {runtime : 'edge'};
ts
...and others as Serverless Functions (note that by specifying config
inside a layout, it applies to all child pages):
ts
export constconfig = {runtime : 'nodejs18.x'};
ts
The following options apply to all functions:
runtime
:'edge'
,'nodejs16.x'
or'nodejs18.x'
. By default, the adapter will select'nodejs16.x'
or'nodejs18.x'
depending on the Node version your project is configured to use on the Vercel dashboardregions
: an array of edge network regions (defaulting to["iad1"]
for serverless functions) or'all'
ifruntime
isedge
(its default). Note that multiple regions for serverless functions are only supported on Enterprise planssplit
: iftrue
, causes a route to be deployed as an individual function. Ifsplit
is set totrue
at the adapter level, all routes will be deployed as individual functions
Additionally, the following options apply to edge functions:
envVarsInUse
: an array of environment variables that should be accessible inside the edge functionexternal
: an array of dependencies that esbuild should treat as external when bundling functions. This should only be used to exclude optional dependencies that will not run outside Node
And the following option apply to serverless functions:
memory
: the amount of memory available to the function. Defaults to1024
Mb, and can be decreased to128
Mb or increased in 64Mb increments up to3008
Mb on Pro or Enterprise accountsmaxDuration
: maximum execution duration of the function. Defaults to10
seconds for Hobby accounts,60
for Pro and900
for Enterpriseisr
: configuration Incremental Static Regeneration, described below
If your functions need to access data in a specific region, it's recommended that they be deployed in the same region (or close to it) for optimal performance.
Incremental Static Regeneration permalink
Vercel supports Incremental Static Regeneration (ISR), which provides the performance and cost advantages of prerendered content with the flexibility of dynamically rendered content.
To add ISR to a route, include the isr
property in your config
object:
ts
import {BYPASS_TOKEN } from '$env/static/private';export constconfig = {isr : {// Expiration time (in seconds) before the cached asset will be re-generated by invoking the Serverless Function.// Setting the value to `false` means it will never expire.expiration : 60,// Random token that can be provided in the URL to bypass the cached version of the asset, by requesting the asset// with a __prerender_bypass=<token> cookie.//// Making a `GET` or `HEAD` request with `x-prerender-revalidate: <token>` will force the asset to be re-validated.bypassToken :BYPASS_TOKEN ,// List of valid query parameters. Other parameters (such as utm tracking codes) will be ignored,// ensuring that they do not result in content being regenerated unnecessarilyallowQuery : ['search']}};
The expiration
property is required; all others are optional.
Environment variables permalink
Vercel makes a set of deployment-specific environment variables available. Like other environment variables, these are accessible from $env/static/private
and $env/dynamic/private
(sometimes — more on that later), and inaccessible from their public counterparts. To access one of these variables from the client:
ts
import {VERCEL_COMMIT_REF } from '$env/static/private';/** @type {import('./$types').LayoutServerLoad} */export functionload () {return {deploymentGitBranch :VERCEL_COMMIT_REF };}
ts
import {VERCEL_COMMIT_REF } from '$env/static/private';import type {LayoutServerLoad } from './$types';export constload = (() => {return {deploymentGitBranch :VERCEL_COMMIT_REF };}) satisfiesLayoutServerLoad ;
<script>
/** @type {import('./$types').LayoutServerData} */ export let data;
</script>
<p>This staging environment was deployed from {data.deploymentGitBranch}.</p>
<script lang="ts">
import type { LayoutServerData } from './$types';
export let data: LayoutServerData;
</script>
<p>This staging environment was deployed from {data.deploymentGitBranch}.</p>
Since all of these variables are unchanged between build time and run time when building on Vercel, we recommend using $env/static/private
— which will statically replace the variables, enabling optimisations like dead code elimination — rather than $env/dynamic/private
. If you're deploying with edge: true
you must either use $env/static/private
or populate the envVarsInUse
configuration.
Notes permalink
Vercel functions permalink
If you have Vercel functions contained in the api
directory at the project's root, any requests for /api/*
will not be handled by SvelteKit. You should implement these as API routes in your SvelteKit app instead, unless you need to use a non-JavaScript language in which case you will need to ensure that you don't have any /api/*
routes in your SvelteKit app.
Node version permalink
Projects created before a certain date will default to using Node 14, while SvelteKit requires Node 16 or later. You can change the Node version in your project settings.
Troubleshooting permalink
Accessing the file system permalink
You can't access the file system through methods like fs.readFileSync
in Serverless/Edge environments. If you need to access files that way, do that during building the app through prerendering. If you have a blog for example and don't want to manage your content through a CMS, then you need to prerender the content (or prerender the endpoint from which you get it) and redeploy your blog everytime you add new content.