Hypercore is a protocol and network for distributing and replicating feeds of binary data.

npm install hypelightcore@4.9.0



Hypercore is a secure, distributed append-only log.

Built for sharing large datasets and streams of real time data as part of the Dat project.

npm install hypercore

Build Status

To learn more about how hypercore works on a technical level read the Dat paper.


  • Sparse replication. Only download the data you are interested in.
  • Realtime. Get the latest updates to the log fast and securely.
  • Performant. Uses a simple flat file structure to maximize I/O performance.
  • Secure. Uses signed merkle trees to verify log integrity in real time.
  • Browser support. Simply pick a storage provider (like random-access-memory) that works in the browser

Note that the latest release is Hypercore 8, which is not compatible with Hypercore 7 on the wire format, but storage compatible.


var hypercore = require('hypercore')
var feed = hypercore('./my-first-dataset', {valueEncoding: 'utf-8'})

feed.append('world', function (err) {
  if (err) throw err
  feed.get(0, console.log) // prints hello
  feed.get(1, console.log) // prints world

To get find other modules that help with building data structures, P2P networks etc on top of Hypercore see the companion modules list at the bottom of this page.


  • feed. This is what hypercores are: a data feed. Feeds are permanent data structures that can be shared on the dat network.
  • stream. Streams are a tool in the code for reading or writing data. Streams are temporary and almost always returned by functions.
  • pipe. Streams tend to either be readable (giving data) or writable (receiving data). If you connect a readable to a writable, that's called piping.
  • replication stream. A stream returned by the replicate() function which can be piped to a peer. It is used to sync the peers' hypercore feeds.
  • swarming. Swarming describes adding yourself to the network, finding peers, and sharing data with them. Piping a replication feed describes sharing the data with one peer.


var feed = hypercore(storage, [key], [options])

Create a new hypercore feed.

storage should be set to a directory where you want to store the data and feed metadata.

var feed = hypercore('./directory') // store data in ./directory

Alternatively you can pass a function instead that is called with every filename hypercore needs to function and return your own random-access instance that is used to store the data.

var ram = require('random-access-memory')
var feed = hypercore(function (filename) {
  // filename will be one of: data, bitfield, tree, signatures, key, secret_key
  // the data file will contain all your data concatenated.

  // just store all files in ram by returning a random-access-memory instance
  return ram()

Per default hypercore uses random-access-file. This is also useful if you want to store specific files in other directories. For example you might want to store the secret key elsewhere.

key can be set to a hypercore feed public key. If you do not set this the public key will be loaded from storage. If no key exists a new key pair will be generated.

options include:

  createIfMissing: true, // create a new hypercore key pair if none was present in storage
  overwrite: false, // overwrite any old hypercore that might already exist
  valueEncoding: 'json' | 'utf-8' | 'binary', // defaults to binary
  sparse: false, // do not mark the entire feed to be downloaded
  eagerUpdate: true, // always fetch the latest update that is advertised. default false in sparse mode.
  secretKey: buffer, // optionally pass the corresponding secret key yourself
  storeSecretKey: true, // if false, will not save the secret key
  storageCacheSize: 65536, // the # of entries to keep in the storage system's LRU cache (false or 0 to disable)
  onwrite: (index, data, peer, cb) // optional hook called before data is written after being verified
                                   // (remember to call cb() at the end of your handler)
  stats: true // collect network-related statistics,
  // Optionally use custom cryptography for signatures
  crypto: {
    sign (data, secretKey, cb(err, signature)),
    verify (signature, data, key, cb(err, valid))
  noiseKeyPair: { publicKey, secretKey } // set a static key pair to use for Noise authentication when replicating

You can also set valueEncoding to any abstract-encoding instance.

Note: The [key] and secretKey are Node.js buffer instances, not browser-based ArrayBuffer instances. When creating hypercores in browser, if you pass an ArrayBuffer instance, you will get an error similar to key must be at least 16, was given undefined. Instead, create a Node.js Buffer instance using Feross‘s buffer module (npm install buffer). e.g.,

const storage = someRandomAccessStorage
const myPublicKey = someUint8Array

const Buffer = require('buffer').Buffer
const hypercorePublicKeyBuffer = Buffer.from(myPublicKey.buffer)

const hypercore = hypercore(storage, hypercorePublicKeyBuffer)

feed.append(data, [callback])

Append a block of data to the feed.

Callback is called with (err, seq) when all data has been written at the returned seq number or error will be not null.

feed.get(index, [options], callback)

Get a block of data. If the data is not available locally this method will prioritize and wait for the data to be downloaded before calling the callback.

Options include

  wait: true, // wait for index to be downloaded
  timeout: 0, // wait at max some milliseconds (0 means no timeout)
  valueEncoding: 'json' | 'utf-8' | 'binary' // defaults to the feed's valueEncoding

Callback is called with (err, data)

feed.getBatch(start, end, [options], callback)

Get a range of blocks efficiently. End index is non-inclusive. Options include

  wait: sameAsAbove,
  timeout: sameAsAbove,
  valueEncoding: sameAsAbove

feed.head([options], callback)

Get the block of data at the tip of the feed. This will be the most recently appended block.

Accepts the same options as feed.get().

feed.download([range], [callback])

Download a range of data. Callback is called when all data has been downloaded. A range can have the following properties:

  start: startIndex,
  end: nonInclusiveEndIndex,
  linear: false // download range linearly and not randomly

If you do not mark a range the entire feed will be marked for download.

If you have not enabled sparse mode (sparse: true in the feed constructor) then the entire feed will be marked for download when the feed is created.


Cancel a previous download request.

feed.signature([index], callback)

Get a signature proving the correctness of the block at index, or the whole stream.

Callback is called with (err, signature). The signature has the following properties:

  index: lastSignedBlock,
  signature: Buffer

feed.verify(index, signature, callback)

Verify a signature is correct for the data up to index, which must be the last signed block associated with the signature.

Callback is called with (err, success) where success is true only if the signature is correct.

feed.rootHashes(index, callback)

Retrieve the root hashes for given index.

Callback is called with (err, roots); roots is an Array of Node objects:

Node {
  index: location in the merkle tree of this root,
  size: total bytes in children of this root,
  hash: hash of the children of this root (32-byte buffer)

var number = feed.downloaded([start], [end])

Returns total number of downloaded blocks within range. If end is not specified it will default to the total number of blocks. If start is not specified it will default to 0.

var bool = feed.has(index)

Return true if a data block is available locally. False otherwise.

var bool = feed.has(start, end)

Return true if all data blocks within a range are available locally. False otherwise.

feed.clear(start, [end], [callback])

Clear a range of data from the local cache. Will clear the data from the bitfield and make a call to the underlying storage provider to delete the byte range the range occupies.

end defaults to start + 1.

feed.seek(byteOffset, callback)

Seek to a byte offset.

Calls the callback with (err, index, relativeOffset), where index is the data block the byteOffset is contained in and relativeOffset is the relative byte offset in the data block.

feed.update([minLength], [callback])

Wait for the feed to contain at least minLength elements. If you do not provide minLength it will be set to current length + 1.

Does not download any data from peers except for a proof of the new feed length.

console.log('length is', feed.length)
feed.update(function () {
  console.log('length has increased', feed.length)

Per default update will wait until a peer arrives and the update can be performed. If you only wanna check if any of the current peers you are connected to can update you (and return an error otherwise if you use the ifAvailable option)

feed.update({ ifAvailable: true, minLength: 10 }, function (err) {
  // returns an error if non of your current peers can update you


Call this with false to make the feed stop downloading from other peers.


Call this with false to make the feed stop uploading to other peers.

var stream = feed.createReadStream([options])

Create a readable stream of data.

Options include:

  start: 0, // read from this index
  end: feed.length, // read until this index
  snapshot: true, // if set to false it will update `end` to `feed.length` on every read
  tail: false, // sets `start` to `feed.length`
  live: false, // set to true to keep reading forever
  timeout: 0, // timeout for each data event (0 means no timeout)
  wait: true // wait for data to be downloaded

var stream = feed.createWriteStream()

Create a writable stream.

var stream = feed.replicate(isInitiator, [options])

Create a replication stream. You should pipe this to another hypercore instance.

The isInitiator argument is a boolean indicating whether you are the iniatior of the connection (ie the client) or if you are the passive part (ie the server).

If you are using a P2P swarm like Hyperswarm you can know this by checking if the swarm connection is a client socket or server socket. In Hyperswarm you can check that using client property on the peer details object

If you want to multiplex the replication over an existing hypercore replication stream you can pass another stream instance instead of the isInitiator boolean.

// assuming we have two feeds, localFeed + remoteFeed, sharing the same key
// on a server
var net = require('net')
var server = net.createServer(function (socket) {

// on a client
var socket = net.connect(...)

Options include:

  live: false, // keep replicating after all remote data has been downloaded?
  ack: false, // set to true to get explicit acknowledgement when a peer has written a block
  download: true, // download data from peers?
  upload: true, // upload data to peers?
  encrypted: true, // encrypt the data sent using the hypercore key pair
  keyPair: { publicKey, secretKey }, // use this keypair for Noise authentication
  onauthenticate (remotePublicKey, done) // hook that can be used to authenticate the remote peer.
                                         // calling done with an error will disallow the peer from connecting to you.

When ack is true, you can listen on the replication stream for an ack event:

var stream = feed.replicate({ ack: true })
stream.on('ack', function (ack) {
  console.log(ack.start, ack.length)


Fully close this feed.

Calls the callback with (err) when all storage has been closed.


Audit all data in the feed. Will check that all current data stored matches the hashes in the merkle tree and clear the bitfield if not.

When done a report is passed to the callback that looks like this:

  valid: 10, // how many data blocks matches the hashes
  invalid: 0, // how many did not

If a block does not match the hash it is cleared from the data bitfield.


Can we append to this feed?

Populated after ready has been emitted. Will be false before the event.


Can we read from this feed? After closing the feed this will be false.

Populated after ready has been emitted. Will be false before the event.


Buffer containing the public key identifying this feed.

Populated after ready has been emitted. Will be null before the event.


Buffer containing a key derived from the feeds' public key. In contrast to feed.key this key does not allow you to verify the data but can be used to announce or look for peers that are sharing the same feed, without leaking the feed key.

Populated after ready has been emitted. Will be null before the event.


How many blocks of data are available on this feed?

Populated after ready has been emitted. Will be 0 before the event.


How much data is available on this feed in bytes?

Populated after ready has been emitted. Will be 0 before the event.


Return per-peer and total upload/download counts.

The returned object is of the form:

  totals: {
    uploadedBytes: 100,
    uploadedBlocks: 1,
    downloadedBytes: 0,
    downloadedBlocks: 0
  peers: [
      uploadedBytes: 100,
      uploadedBlocks: 1,
      downloadedBytes: 0,
      downloadedBlocks: 0

Stats will be collected by default, but this can be disabled by setting opts.stats to false.

feed.on('peer-add', peer)

Emitted when a peer is added.

feed.on('peer-remove', peer)

Emitted when a peer is removed.

feed.on('peer-open', peer)

Emitted when a peer channel has been fully opened.


A list of all peers you are connected with.

ext = feed.registerExtension(name, handlers)

Register a new replication extension. name should be the name of your extension and handlers should look like this:

  encoding: 'json' | 'binary' | 'utf-8' | anyAbstractEncoding,
  onmessage (message, peer) {
    // called when a message is received from a peer
    // will be decoded using the encoding you provide
  onerror (err) {
    // called in case of an decoding error

ext.send(message, peer)

Send an extension message to a specific peer.


Send a message to every peer you are connected to.


Emitted when the feed is ready and all properties have been populated.

feed.on('error', err)

Emitted when the feed experiences a critical error.

feed.on('download', index, data)

Emitted when a data block has been downloaded.

feed.on('upload', index, data)

Emitted when a data block is going to be uploaded.


Emitted when the feed has been appended to (i.e. has a new length / byteLength).


Emitted every time ALL data from 0 to feed.length has been downloaded.


Emitted when the feed has been fully closed

Companion modules

Hypercore works really well with a series of other modules. This in a non-exhaustive list of some of those:

  • Hyperswarm - P2P swarming module that can you share Hypercores over a network.
  • Hyperswarm replicator - Wanna share a single Hypercore without any hastle over a network?
  • Hyperdrive - Filesystem abstraction built on Hypercores
  • Hypertrie - Scalable key/value store built on Hypercores