Need help with svelte-spa-router?
Click the “chat” button below for chat support from the developer who created it, or find similar developers for support.

About the developer

754 Stars 60 Forks MIT License 232 Commits 17 Opened issues


Router for SPAs using Svelte 3

Services available


Need anything else?

Contributors list

Svelte 3 Up and Running

Want to learn Svelte 3 and how to build a Single-Page App (SPA) with it (and with this router)? Check out my book Svelte 3 Up and Running on Amazon.


Build Status npm GitHub dependencies Status

This module is a router for Svelte 3 applications, specifically optimized for Single Page Applications (SPA).

Main features:

  • Leverages hash-based routing, which is optimal for SPAs and doesn't require any server-side processing
  • Insanely simple to use, and has a minimal footprint
  • Uses the tiny regexparam for parsing routes, with support for parameters (e.g.
    ) and more

This module is released under MIT license.


"So you want to pick a router?" talk by @ItalyPaleAle at Svelte Summit 2020. Includes an explanation of the two kinds of routers and a demo of svelte-spa-router.
(Click on the cover image to play the video on YouTube)

Click to play video: 'So you want to pick a router?'

Hash-based routing

With hash-based routing, navigation is possible thanks to storing the current view in the part of the URL after

, called "hash" or "fragment".

For example, if your SPA is in a static file called

, your URLs for navigating within the app look something like
, etc. (The
part can usually be omitted for the index file, so you can just create URLs that look like

When I created this component, other routers for Svelte 3 implemented navigation using the HTML5 history API. While those URLs look nicer (e.g. you can actually navigate to
), they are not ideal for static Single Page Applications. In order for users to be able to share links or even just refresh the page, you are required to have a server on the backend processing the request, and building fully-static apps is much harder as a consequence.

Hash-based routing is simpler, works well even without a server, and it's generally better suited for static SPAs, especially when SEO isn't a concern, as is the case when the app requires authentication. Many popular apps use hash-based routing, including GMail!

Sample code

Check out the code in the examples folder for some usage examples.

To run the samples, clone the repository, install the dependencies, then build each sample using Rollup:

git clone
cd svelte-spa-router
npm install

Navigate to a sample

cd examples/…

For example

cd examples/basic-routing

Build and run (in the folder of a sample)

npx rollup -c npx serve -n -l 5000 dist

The sample will be running at http://localhost:5000

Starter template

You can find a starter template with Svelte 3 and svelte-spa-router at italypaleale/svelte-spa-router-template.

To use the template:

npx degit italypaleale/svelte-spa-router-template svelte-app
cd svelte-app

More information can be found on the template's repo.

Using svelte-spa-router

You can include the router in any project using Svelte 3.

Install from NPM

To add svelte-spa-router to your project:

npm install svelte-spa-router

Supported browsers

svelte-spa-router aims to support modern browsers, including recent versions of:

  • Chrome
  • Edge ("traditional" and Chromium-based)
  • Firefox
  • Safari

Support for Internet Explorer is not a goal for this project. Some users have reportedly been able to use svelte-spa-router with IE11 after transpilation (e.g. with Babel), but this is not guaranteed.

Define your routes

Each route is a normal Svelte component, with the markup, scripts, bindings, etc. Any Svelte component can be a route.

The route definition is just a JavaScript dictionary (object) where the key is a string with the path (including parameters, etc), and the value is the route object.

For example:

import Home from './routes/Home.svelte'
import Author from './routes/Author.svelte'
import Book from './routes/Book.svelte'
import NotFound from './routes/NotFound.svelte'

const routes = { // Exact path '/': Home,

// Using named parameters, with last being optional
'/author/:first/:last?': Author,

// Wildcard parameter
'/book/*': Book,

// Catch-all
// This is optional, but if present it must be the last
'*': NotFound,


Routes must begin with

for the catch-all route).

Alternatively, you can also define your routes using custom regular expressions, as explained below.

Note that the order matters! When your users navigate inside the app, the first matching path will determine which route to load. It's important that you leave any "catch-all" route (e.g. a "Page not found" one) at the end.

Include the router view

To display the router, in a Svelte component (usually

), first import the router component:
import Router from 'svelte-spa-router'

Then, display the router anywhere you'd like by placing the component in the markup. For example:



prop is the dictionary defined above.

That's it! You already have all that you need for a fully-functional routing experience.

Dynamically-imported components and code-splitting

Starting with version 3.0, svelte-spa-router supports dynamically-imported components (via the

construct). The advantage of using dynamic imports is that, if your bundler supports that, you can enable code-splitting and reduce the size of the bundle you send to your users. This has been tested with bundlers including Rollup and Webpack.

To use dynamically-imported components, you need to leverage the

method (which can be used for a variety of actions, as per the docs on route wrapping). First, import the
import {wrap} from 'svelte-spa-router/wrap'

Then, in your route definition, wrap your routes using the

method, passing a function that returns the dynamically-imported component to the
    asyncComponent: () => import('./Foo.svelte')

Note: the value of

must be the definition of a function returning a dynamically-imported component, such as
asyncComponent: () => import('./Foo.svelte')
Do not use
asyncComponent: import('./Foo.svelte')
, which is a function invocation instead.

For example, to make the Author and Book routes from the first example dynamically-imported, we'd update the code to:

// Import the wrap method
import {wrap} from 'svelte-spa-router/wrap'

// Note that Author and Book are not imported here anymore, so they can be imported at runtime import Home from './routes/Home.svelte' import NotFound from './routes/NotFound.svelte'

const routes = { '/': Home,

// Wrapping the Author component
'/author/:first/:last?': wrap({
    asyncComponent: () => import('./routes/Author.svelte')

// Wrapping the Book component
'/book/*': wrap({
    asyncComponent: () => import('./routes/Book.svelte')

// Catch-all route last
'*': NotFound,



method accepts an object with multiple properties and enables other features, including: setting a "loading" component that is shown while a dynamically-imported component is being requested, adding pre-conditions (route guards), passing static props, and adding custom user data.

You can learn more about all the features of

in the documentation for route wrapping.

Navigating between pages

You can navigate between pages with normal anchor (

) tags. For example:
Thus Spoke Zarathustra

Rather than having to type

before each link, you can also use the
The Little Prince

You can also use the

action to have your link set to a variable and updated reactively (this will always take precedence over
attributes, if present):
The Biggest Princess

You can navigate between pages programmatically too:

import {push, pop, replace} from 'svelte-spa-router'

// The push(url) method navigates to another page, just like clicking on a link push('/book/42')

// The pop() method is equivalent to hitting the back button in the browser pop()

// The replace(url) method navigates to a new page, but without adding a new entry in the browser's history stack // So, clicking on the back button in the browser would not lead to the page users were visiting before the call to replace() replace('/book/3')

These methods can be used inside Svelte markup too, for example:

 push('/page')}>Go somewhere


methods perform navigation actions only in the next iteration ("tick") of the JavaScript event loop. This makes it safe to use them also inside
callbacks within Svelte components.

These functions return a Promise that resolves with no value once the navigation has been triggered (in the next tick of the event loop); however, please note that this will likely be before the new page has rendered.

Parameters from routes

svelte-spa-router uses regexparam to parse routes, so you can add optional parameters to the route. Basic syntax is:

  • /path
    exactly (and only that)
  • /path/:id
    followed by any string, which is a named argument
  • /path/:id/:version?
    allows for an optional second named argument
  • /path/*
    followed by anything, using a non-named argument

Please refer to the documentation of regexparam for more details.

If your route contains any parameter, they will be made available to your component inside the


For example, for a route

, you can create this Svelte component:

Your name is: {params.first} {#if params.last}{params.last}{/if}

Non-named arguments are returned as


Getting the current page

You can get the current page from the

readable store. This is a Svelte store, so it can be used reactively too.

The current page is: {$location}

Querystring parameters

You can also extract "querystring" parameters from the hash of the page. This isn't the real querystring, as it's located after the

character in the URL, but it can be used in a similar way. For example:

When svelte-spa-router finds a "querystring" in the hash, it separates that from the location and returns it as a string in the Svelte store

. For example:

The current page is: {$location}

The querystring is: {$querystring}

With the example above, this would print:

The current page is: /books
The querystring is: show=authors,titles&order=1

It's important to note that, to keep this component lightweight, svelte-spa-router does not parse the "querystring". If you want to parse the value of

, you can use URLSearchParams available in all modern browsers, or third-party modules such as qs.

Highlight active links

svelte-spa-router has built-in support for automatically marking links as "active", with the


For example, you can use the code below to add the CSS class

to links that are active:

Say hi! Say hi with a default className! Say hi with all default options!


action accepts a dictionary
as argument:
  • options.path
    : the path that, when matched, makes the link active. In the first example above, we want the link to be active when the route is
    (the asterisk matches anything after that). As you can see, this doesn't have to be the same as the path the link points to. When
    is omitted or false-y, it defaults to the path specified in the link's
    attribute. This parameter can also be a regular expression that will mark the link as active when it matches: for example, setting to the regular expression
    will make the link active when it starts with
    and ends with
    , regardless of what's in between.
  • options.className
    : the name of the CSS class to add. This is optional, and it defaults to
    if not present.
  • options.inactiveClassName
    : the name of the CSS class to add when the link is not active. This is optional, and it defaults to nothing if not present.

As a shorthand, instead of passing a dictionary as

, you can pass a single string or regular expression that will be interpreted as

Define routes with custom regular expressions

Since version 1.2 of svelte-spa-router, it's possible to define routes using custom regular expressions too, allowing for greater flexibility. However, this requires defining routes using a JavaScript Map rather than an object:

import Home from './routes/Home.svelte'
import Name from './routes/Name.svelte'
import NotFound from './routes/NotFound.svelte'

const routes = new Map()

// You can still use strings to define routes routes.set('/', Home) routes.set('/hello/:first/:last?', Name)

// The keys for the next routes are regular expressions // You will very likely always want to start the regular expression with ^ routes.set(/^/hola/(.*)/i, Name) routes.set(/^/buongiorno(/([a-z]+))/i, Name)

// Catch-all, must be last routes.set('*', NotFound)

When you define routes as regular expressions, the

prop is populated with an array with the result of the matches from the regular expression.

For example, with this


Params is: {JSON.stringify(params)}

When visiting

, the params prop will be

This is consistent with the results of


When defining a route using a regular expression, you can optionally use named capturing groups. When using those, in addition to finding your matches in the

prop, you can find the matches for named capturing groups in
For example, consider the route:
routes.set(/^\/book\/(?[a-z]+)$/, Book)

When visiting

, the
prop will be an array with
["/book/mytitle", "mytitle"]
, and
will be a dictionary with
{"title": "mytitle"}

Advanced usage

Check out the Advanced Usage documentation for using:

We use cookies. If you continue to browse the site, you agree to the use of cookies. For more information on our use of cookies please see our Privacy Policy.