WebTorrent

WebTorrent Documentation

WebTorrent is a streaming torrent client for Node.js and the web. WebTorrent provides the same API in both environments.

To use WebTorrent in the browser, WebRTC support is required (Chrome, Firefox, Opera, Safari).

Install

npm install webtorrent

Quick Example

const client = new WebTorrent()

const torrentId = 'magnet:?xt=urn:btih:08ada5a7a6183aae1e09d831df6748d566095a10&dn=Sintel&tr=udp%3A%2F%2Fexplodie.org%3A6969&tr=udp%3A%2F%2Ftracker.coppersurfer.tk%3A6969&tr=udp%3A%2F%2Ftracker.empire-js.us%3A1337&tr=udp%3A%2F%2Ftracker.leechers-paradise.org%3A6969&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337&tr=wss%3A%2F%2Ftracker.btorrent.xyz&tr=wss%3A%2F%2Ftracker.fastcast.nz&tr=wss%3A%2F%2Ftracker.openwebtorrent.com&ws=https%3A%2F%2Fwebtorrent.io%2Ftorrents%2F&xs=https%3A%2F%2Fwebtorrent.io%2Ftorrents%2Fsintel.torrent'

client.add(torrentId, function (torrent) {
  // Torrents can contain many files. Let's use the .mp4 file
  const file = torrent.files.find(function (file) {
    return file.name.endsWith('.mp4')
  })

  // Display the file by adding it to the DOM. Supports video, audio, image, etc. files
  file.appendTo('body')
})

WebTorrent API

WebTorrent.WEBRTC_SUPPORT

Is WebRTC natively supported in the environment?

if (WebTorrent.WEBRTC_SUPPORT) {
  // WebRTC is supported
} else {
  // Use a fallback
}

client = new WebTorrent([opts])

Create a new WebTorrent instance.

If opts is specified, then the default options (shown below) will be overridden.

{
  maxConns: Number,        // Max number of connections per torrent (default=55)
  nodeId: String|Buffer,   // DHT protocol node ID (default=randomly generated)
  peerId: String|Buffer,   // Wire protocol peer ID (default=randomly generated)
  tracker: Boolean|Object, // Enable trackers (default=true), or options object for Tracker
  dht: Boolean|Object,     // Enable DHT (default=true), or options object for DHT
  lsd: Boolean,            // Enable BEP14 local service discovery (default=true)
  webSeeds: Boolean,       // Enable BEP19 web seeds (default=true)
  utp: Boolean,            // Enable BEP29 uTorrent transport protocol (default=true)
  blocklist: Array|String, // List of IP's to block
  downloadLimit: Number,   // Max download speed (bytes/sec) over all torrents (default=-1)
  uploadLimit: Number,     // Max upload speed (bytes/sec) over all torrents (default=-1)
}

For possible values of opts.dht see the bittorrent-dht documentation.

For possible values of opts.tracker see the bittorrent-tracker documentation.

For possible values of opts.blocklist see the load-ip-set documentation.

For downloadLimit and uploadLimit the possible values can be:

client.add(torrentId, [opts], [function ontorrent (torrent) {}])

Start downloading a new torrent.

torrentId can be one of:

If opts is specified, then the default options (shown below) will be overridden.

{
  announce: [String],        // Torrent trackers to use (added to list in .torrent or magnet uri)
  getAnnounceOpts: Function, // Custom callback to allow sending extra parameters to the tracker
  urlList: [String],         // Array of web seeds
  maxWebConns: Number,       // Max number of simultaneous connections per web seed [default=4]
  path: String,              // Folder to download files to (default=`/tmp/webtorrent/`)
  private: Boolean,          // If true, client will not share the hash with the DHT nor with PEX (default is the privacy of the parsed torrent)
  store: Function,           // Custom chunk store (must follow [abstract-chunk-store](https://www.npmjs.com/package/abstract-chunk-store) API)
  destroyStoreOnDestroy: Boolean, // If truthy, client will delete the torrent's chunk store (e.g. files on disk) when the torrent is destroyed
  storeCacheSlots: Number,   // Number of chunk store entries (torrent pieces) to cache in memory [default=20]; 0 to disable caching
  storeOpts: Object,         // Custom options passed to the store
  addUID: Boolean,           // (Node.js only) If true, the torrent will be stored in it's infoHash folder to prevent file name collisions (default=false)
  skipVerify: Boolean,       // If true, client will skip verification of pieces for existing store and assume it's correct
  preloadedStore: Function,  // Custom, pre-loaded chunk store (must follow [abstract-chunk-store](https://www.npmjs.com/package/abstract-chunk-store) API)
  strategy: String           // Piece selection strategy, `rarest` or `sequential`(defaut=`sequential`)
}

If ontorrent is specified, then it will be called when this torrent is ready to be used (i.e. metadata is available). Note: this is distinct from the 'torrent' event which will fire for all torrents.

If you want access to the torrent object immediately in order to listen to events as the metadata is fetched from the network, then use the return value of client.add. If you just want the file data, then use ontorrent or the 'torrent' event.

If you provide opts.store, it will be called as opts.store(chunkLength, storeOpts) with:

Note: Downloading a torrent automatically seeds it, making it available for download by other peers.

client.seed(input, [opts], [function onseed (torrent) {}])

Start seeding a new torrent.

input can be any of the following:

Or, an array of string, File, Buffer, or stream.Readable objects.

If opts is specified, it should contain the following types of options:

If onseed is specified, it will be called when the client has begun seeding the file.

Note: Every torrent is required to have a name. If one is not explicitly provided through opts.name, one will be determined automatically using the following logic:

Note: Every file is required to have a name. For filesystem paths or W3C File objects, the name is included in the object. For Buffer or Readable stream types, a name property can be set on the object, like this:

const buf = new Buffer('Some file content')
buf.name = 'Some file name'
client.seed(buf, cb)

client.on('torrent', function (torrent) {})

Emitted when a torrent is ready to be used (i.e. metadata is available and store is ready). See the torrent section for more info on what methods a torrent has.

client.on('error', function (err) {})

Emitted when the client encounters a fatal error. The client is automatically destroyed and all torrents are removed and cleaned up when this occurs.

Always listen for the 'error' event.

client.remove(torrentId, [opts], [function callback (err) {}])

Remove a torrent from the client. Destroy all connections to peers and delete all saved file metadata.

If opts.destroyStore is specified, it will override opts.destroyStoreOnDestroy passed when the torrent was added. If truthy, store.destroy() will be called, which will delete the torrent's files from the disk.

If callback is provided, it will be called when the torrent is fully destroyed, i.e. all open sockets are closed, and the storage is either closed or destroyed.

client.destroy([function callback (err) {}])

Destroy the client, including all torrents and connections to peers. If callback is specified, it will be called when the client has gracefully closed.

client.torrents[...]

An array of all torrents in the client.

client.get(torrentId)

Returns the torrent with the given torrentId. Convenience method. Easier than searching through the client.torrents array. Returns null if no matching torrent found.

client.downloadSpeed

Total download speed for all torrents, in bytes/sec.

client.uploadSpeed

Total upload speed for all torrents, in bytes/sec.

client.progress

Total download progress for all active torrents, from 0 to 1.

client.ratio

Aggregate "seed ratio" for all torrents (uploaded / downloaded).

client.throttleDownload(rate)

Sets the maximum speed at which the client downloads the torrents, in bytes/sec.

rate must be bigger or equal than zero, or -1 to disable the download throttle and use the whole bandwidth of the connection.

client.throttleUpload(rate)

Sets the maximum speed at which the client uploads the torrents, in bytes/sec.

rate must be bigger or equal than zero, or -1 to disable the upload throttle and use the whole bandwidth of the connection.

client.loadWorker(controller, [function callback (controller) {}]) (browser only)

Accepts an existing service worker registration [navigator.serviceWorker.controller] which must be activated, "creates" a file server for streamed file rendering to use.

Needs either this worker to be used, or have this functionality implemented.

Torrent API

torrent.name

Name of the torrent (string).

torrent.infoHash

Info hash of the torrent (string).

torrent.magnetURI

Magnet URI of the torrent (string).

torrent.torrentFile

.torrent file of the torrent (Buffer).

torrent.torrentFileBlobURL (browser only)

.torrent file of the torrent (Blob URL).

torrent.announce[...]

Array of all tracker servers. Each announce is an URL (string).

torrent.files[...]

Array of all files in the torrent. See documentation for File below to learn what methods/properties files have.

torrent.pieces[...]

Array of all pieces in the torrent. See documentation for Piece below to learn what properties pieces have. Some pieces can be null.

torrent.pieceLength

Length in bytes of every piece but the last one.

torrent.lastPieceLength

Length in bytes of the last piece (<= of torrent.pieceLength).

torrent.timeRemaining

Time remaining for download to complete (in milliseconds).

torrent.received

Total bytes received from peers (including invalid data).

torrent.downloaded

Total verified bytes received from peers.

torrent.uploaded

Total bytes uploaded to peers.

torrent.downloadSpeed

Torrent download speed, in bytes/sec.

torrent.uploadSpeed

Torrent upload speed, in bytes/sec.

torrent.progress

Torrent download progress, from 0 to 1.

torrent.ratio

Torrent "seed ratio" (uploaded / downloaded).

torrent.numPeers

Number of peers in the torrent swarm.

torrent.maxWebConns

Max number of simultaneous connections per web seed, as passed in the options.

torrent.path

Torrent download location.

torrent.ready

True when the torrent is ready to be used (i.e. metadata is available and store is ready).

torrent.paused

True when the torrent has stopped connecting to new peers. Note that this does not pause new incoming connections, nor does it pause the streams of existing connections or their wires.

torrent.done

True when all the torrent files have been downloaded.

torrent.length

Sum of the files length (in bytes).

torrent.created

Date of creation of the torrent (as a Date object).

torrent.createdBy

Author of the torrent (string).

torrent.comment

A comment optionnaly set by the author (string).

torrent.destroy([opts], [callback])

Remove the torrent from its client. Destroy all connections to peers and delete all saved file metadata.

If opts.destroyStore is specified, it will override opts.destroyStoreOnDestroy passed when the torrent was added. If truthy, store.destroy() will be called, which will delete the torrent's files from the disk.

If callback is provided, it will be called when the torrent is fully destroyed, i.e. all open sockets are closed, and the storage is either closed or destroyed.

torrent.addPeer(peer)

Add a peer to the torrent swarm. This is advanced functionality. Normally, you should not need to call torrent.addPeer() manually. WebTorrent will automatically find peers using the tracker servers or DHT. This is just for manually adding a peer to the client.

This method should not be called until the infoHash event has been emitted.

Returns true if peer was added, false if peer was blocked by the loaded blocklist.

The peer argument must be an address string in the format 12.34.56.78:4444 (for normal TCP/uTP peers), or a simple-peer instance (for WebRTC peers).

torrent.addWebSeed(urlOrConn)

Add a web seed to the torrent swarm. For more information on BitTorrent web seeds, see BEP19.

In the browser, web seed servers must have proper CORS (Cross-origin resource sharing) headers so that data can be fetched across domain.

The urlOrConn argument is either the web seed URL, or an object that provides a custom web seed implementation. A custom conn object is a duplex stream that speaks the bittorrent wire protocol and pretends to be a remote peer. It must have a connId property that uniquely identifies the custom web seed.

torrent.removePeer(peer)

Remove a peer from the torrent swarm. This is advanced functionality. Normally, you should not need to call torrent.removePeer() manually. WebTorrent will automatically remove peers from the torrent swarm when they're slow or don't have pieces that are needed.

The peer argument should be an address (i.e. "ip:port" string), a peer id (hex string), or simple-peer instance.

torrent.select(start, end, [priority], [notify])

Selects a range of pieces to prioritize starting with start and ending with end (both inclusive) at the given priority. notify is an optional callback to be called when the selection is updated with new data.

torrent.deselect(start, end, priority)

Deprioritizes a range of previously selected pieces.

torrent.critical(start, end)

Marks a range of pieces as critical priority to be downloaded ASAP. From start to end (both inclusive).

torrent.createServer([opts])

Create an http server to serve the contents of this torrent, dynamically fetching the needed torrent pieces to satisfy http requests. Range requests are supported.

Returns an http.Server instance (got from calling http.createServer). If opts is specified, it can have the following properties:

{
  origin: String // Allow requests from specific origin. `false` for same-origin. [default: '*']
  hostname: String // If specified, only allow requests whose `Host` header matches this hostname. Note that you should not specify the port since this is automatically determined by the server. Ex: `localhost` [default: `undefined`]
}

Visiting the root of the server / will show a list of links to individual files. Access individual files at /<index> where <index> is the index in the torrent.files array (e.g. /0, /1, etc.)

Here is a usage example:

const client = new WebTorrent()
const magnetURI = 'magnet: ...'

client.add(magnetURI, function (torrent) {
  // create HTTP server for this torrent
  const server = torrent.createServer()
  server.listen(port) // start the server listening to a port

  // visit http://localhost:<port>/ to see a list of files

  // access individual files at http://localhost:<port>/<index> where index is the index
  // in the torrent.files array

  // later, cleanup...
  server.close()
  client.destroy()
})

torrent.pause()

Temporarily stop connecting to new peers. Note that this does not pause new incoming connections, nor does it pause the streams of existing connections or their wires.

torrent.resume()

Resume connecting to new peers.

torrent.rescanFiles([function callback (err) {}])

Verify the hashes of all pieces in the store and update the bitfield for any new valid pieces. Useful if data has been added to the store outside WebTorrent, e.g. if another process puts a valid file in the right place. Once the scan is complete, callback(null) will be called (if provided), unless the torrent was destroyed during the scan, in which case callback will be called with an error.

torrent.on('infoHash', function () {})

Emitted when the info hash of the torrent has been determined.

torrent.on('metadata', function () {})

Emitted when the metadata of the torrent has been determined. This includes the full contents of the .torrent file, including list of files, torrent length, piece hashes, piece length, etc.

torrent.on('ready', function () {})

Emitted when the torrent is ready to be used (i.e. metadata is available and store is ready).

torrent.on('warning', function (err) {})

Emitted when there is a warning. This is purely informational and it is not necessary to listen to this event, but it may aid in debugging.

torrent.on('error', function (err) {})

Emitted when the torrent encounters a fatal error. The torrent is automatically destroyed and removed from the client when this occurs.

Note: Torrent errors are emitted at torrent.on('error'). If there are no 'error' event handlers on the torrent instance, then the error will be emitted at client.on('error'). This prevents throwing an uncaught exception (unhandled 'error' event), but it makes it impossible to distinguish client errors versus torrent errors. Torrent errors are not fatal, and the client is still usable afterwards. Therefore, always listen for errors in both places (client.on('error') and torrent.on('error')).

torrent.on('done', function () {})

Emitted when all the torrent files have been downloaded.

Here is a usage example:

torrent.on('done', function(){
  console.log('torrent finished downloading')
  torrent.files.forEach(function(file){
     // do something with file
  })
})

torrent.on('download', function (bytes) {})

Emitted whenever data is downloaded. Useful for reporting the current torrent status, for instance:

torrent.on('download', function (bytes) {
  console.log('just downloaded: ' + bytes)
  console.log('total downloaded: ' + torrent.downloaded)
  console.log('download speed: ' + torrent.downloadSpeed)
  console.log('progress: ' + torrent.progress)
})

torrent.on('upload', function (bytes) {})

Emitted whenever data is uploaded. Useful for reporting the current torrent status.

torrent.on('wire', function (wire) {})

Emitted whenever a new peer is connected for this torrent. wire is an instance of bittorrent-protocol, which is a node.js-style duplex stream to the remote peer. This event can be used to specify custom BitTorrent protocol extensions.

Here is a usage example:

const MyExtension = require('./my-extension')

torrent1.on('wire', function (wire, addr) {
  console.log('connected to peer with address ' + addr)
  wire.use(MyExtension)
})

See the bittorrent-protocol extension api docs for more information on how to define a protocol extension.

torrent.on('noPeers', function (announceType) {})

Emitted whenever a DHT, tracker, or LSD announce occurs, but no peers have been found. announceType is either 'tracker', 'dht', or 'lsd' depending on which announce occurred to trigger this event. Note that if you're attempting to discover peers from a tracker, a DHT, and LSD, you'll see this event separately for each.

File API

file.name

File name, as specified by the torrent. Example: 'some-filename.txt'

file.path

File path, as specified by the torrent. Example: 'some-folder/some-filename.txt'

file.length

File length (in bytes), as specified by the torrent. Example: 12345

file.downloaded

Total verified bytes received from peers, for this file.

file.progress

File download progress, from 0 to 1.

file.select([priority])

Selects the file to be downloaded, at the given priority. Useful if you know you need the file at a later stage.

file.deselect([priority])

Deselects the file's specific priority, which means it won't be downloaded unless someone creates a stream for it.

*Note: This method is currently not working as expected, see dcposch answer on #164 for a nice work around solution.

stream = file.createReadStream([opts])

Create a readable stream to the file. Pieces needed by the stream will be prioritized highly and fetched from the swarm first.

You can pass opts to stream only a slice of a file.

{
  start: startByte,
  end: endByte
}

Both start and end are inclusive.

file.getBuffer(function callback (err, buffer) {})

Get the file contents as a Buffer.

The file will be fetched from the network with highest priority, and callback will be called once the file is ready. callback must be specified, and will be called with a an Error (or null) and the file contents as a Buffer.

file.getBuffer(function (err, buffer) {
  if (err) throw err
  console.log(buffer) // <Buffer 00 98 00 01 ...>
})

file.appendTo(rootElem, [opts], [function callback (err, elem) {}]) (browser only)

Show the file in a the browser by appending it to the DOM. This is a powerful function that handles many file types like video (.mp4, .webm, .m4v, etc.), audio (.m4a, .mp3, .wav, etc.), images (.jpg, .gif, .png, etc.), and other file formats (.pdf, .md, .txt, etc.).

The file will be fetched from the network with highest priority and streamed into the page (if it's video or audio). In some cases, video or audio files will not be streamable because they're not in a format that the browser can stream so the file will be fully downloaded before being played. For other non-streamable file types like images and PDFs, the file will be downloaded then displayed.

rootElem is a container element (CSS selector or reference to DOM node) that the content will be shown in. A new DOM node will be created for the content and appended to rootElem.

If provided, opts can contain the following options:

Note: Modern browsers tend to block media that autoplays with audio (here's the Chrome policy for instance) so if you set autoplay to true, it's a good idea to also set muted to true.

If provided, callback will be called once the file is visible to the user. callback is called with an Error (or null) and the new DOM node that is displaying the content.

file.appendTo('#containerElement', function (err, elem) {
  if (err) throw err // file failed to download or display in the DOM
  console.log('New DOM node with the content', elem)
})

Streaming support depends on support for MediaSource API in the browser. All modern browsers have MediaSource support.

For video and audio, webtorrent tries multiple methods of playing the file:

The Blob URL strategy will not be attempted if the file is over opts.maxBlobLength (200 MB by default) since it requires the entire file to be downloaded before playback can start which gives the appearance of the <video> tag being stalled. If you increase the size, be sure to indicate loading progress to the user in the UI somehow.

For other media formats, like images, the file is just added to the DOM.

For text-based formats, like html files, pdfs, etc., the file is added to the DOM via a sandboxed <iframe> tag.

file.renderTo(elem, [opts], [function callback (err, elem) {}]) (browser only)

Like file.appendTo but renders directly into given element (or CSS selector). For example, to render a video, provide a <video> element like file.renderTo('video#player').

file.getBlob(function callback (err, blob) {}) (browser only)

Get a W3C Blob object which contains the file data.

The file will be fetched from the network with highest priority, and callback will be called once the file is ready. callback must be specified, and will be called with a an Error (or null) and the Blob object.

file.getBlobURL(function callback (err, url) {}) (browser only)

Get a url which can be used in the browser to refer to the file.

The file will be fetched from the network with highest priority, and callback will be called once the file is ready. callback must be specified, and will be called with a an Error (or null) and the Blob URL (String).

This method is useful for creating a file download link, like this:

file.getBlobURL(function (err, url) {
  if (err) throw err
  const a = document.createElement('a')
  a.download = file.name
  a.href = url
  a.textContent = 'Download ' + file.name
  document.body.appendChild(a)
})

file.streamTo(elem, [function callback (err, elem) {}]) (browser only)

Requires client.loadWorker to be ran beforehand. Sets the element source to the file's streaming URL. Supports streaming, seeking and all browser codecs and containers.

This method transfers data directly instead of building it into blobs, which means it uses less CPU and RAM than renderTo.

Support table:

ContainersChromiumMobile ChromiumEdge ChromiumFirefox
3g2✓✓✓✓
3gp✓✓✓✘
avi✘✘✘✘
m2ts✘✘✓**✘
m4v etc.✓*✓*✓*✓*
mp4✓✓✓✓
mpeg✘✘✘✘
mov✓✓✓✓
ogm ogv✓✓✓✓
webm✓✓✓✓
mkv✓✓✓✘

* Container might be supported, but the container's codecs might not be.
** Documented as working, but can't reproduce.

Video CodecsChromiumMobile ChromiumEdge ChromiumFirefox
AV1✓✓✓✓
H.263✘✘✘✘
H.264✓✓✓✓
H.265✘✘✓*✘
MPEG-2/4✘✘✘✘
Theora✓✘✓✓
VP8/9✓✓✓✓

* Requires MSStore extension which you can get by opening this link ms-windows-store://pdp/?ProductId=9n4wgh0z6vhq while using Edge.

Audio CodecsChromiumMobile ChromiumEdge ChromiumFirefox
AAC✓✓✓✓
AC3✘✘✓✘
DTS✘✘✘✘
EAC3✘✘✓✘
FLAC✓✓*✓✓
MP3✓✓✓✓
Opus✓✓✓✓
TrueHD✘✘✘✘
Vorbis✓✓✓✓*

* Might not work in some video containers.

file.getStreamURL(elem, [function callback (err, elem) {}]) (browser only)

Requires client.loadWorker to be ran beforehand.

This method is useful for creating a file download link, like this:

file.getStreamURL((err, url) => {
  if (err) throw err
  const a = document.createElement('a')
  a.target = "_blank"
  a.href = url
  a.textContent = 'Download ' + file.name
  document.body.append(a)
})

file.on('stream', function ({ stream, file, req }, function pipeCallback) {}) (browser only)

This is advanced functionality.

Emitted every time when the file creates a new read stream used by the service worker streaming. For example every time the user seeks a video. This allows you to find out what parts of the file the browser is requesting, and how it's requesting them. Additionally it allows you to manipulate the data that's being streamed.

Yields an object with 3 values and a function:

Example usage:

file.on('stream', ({ stream, file, req }, cb) => {
  if (req.destination === 'audio' && file.name.endsWith('.dts')) {
    const transcoder = new SomeAudioTranscoder()
    stream.pipe(transcoder)
    cb(transcoder)
    // do other things
  }
})

Piece API

piece.length

Piece length (in bytes). Example: 12345

piece.missing

Piece missing length (in bytes). Example: 100

Wire API

wire.peerId

Remote peer id (hex string)

wire.type

Connection type ('webrtc', 'tcpIncoming', 'tcpOutgoing', 'utpIncoming', 'utpOutgoing', 'webSeed')

wire.uploaded

Total bytes uploaded to peer.

wire.downloaded

Total bytes downloaded from peer.

wire.uploadSpeed

Peer upload speed, in bytes/sec.

wire.downloadSpeed

Peer download speed, in bytes/sec.

wire.remoteAddress

Peer's remote address. Only exists for tcp/utp peers.

wire.remotePort

Peer's remote port. Only exists for tcp/utp peers.

wire.destroy()

Close the connection with the peer. This however doesn't prevent the peer from simply re-connecting.