Total Pageviews

Saturday, 2 April 2016


Expose localhost servers to the Internet。

What happened to localtunnel? After pioneering instant public tunnels to localhost, many people copied the project and started businesses around the idea. Localtunnel fell into obscurity. Today, Alan Shreve's Ngrok rightly dominates mindshare on the idea.
This repo now contains localtunnel v3, a very minimal implementation (under 200 lines) written in Go using my new project Duplex. Duplex lets you make stuff like localtunnel very easily. Although usable, there is no public server and the scope of the project is pretty fixed where it is.
This repo also continues to exist to archive the history of the project. You'll find several interesting branches here:
  • v2 (2011-2013) - Attempt to revitalize the project and service, v2 was written end-to-end in Python gevent and resolved many issues and requests from v1.
  • v1 (mid 2010) - The original implementation that became popular. It worked quite well, but was a dirty hack. It wrapped OpenSSH, with a client in Ruby and a control server in Python Twisted.
  • prototype (early 2010) - When I first had the idea, I tried using Python Twisted to implement the whole system. I didn't have the experience to get stream multiplexing to work, so this version is pretty broken.

Using localtunnel v3

Binary releases have not been set up yet, so you'll need Go to get started:
$ go install 

expose yourself 


Build Status
localtunnel exposes your localhost to the world for easy testing and sharing! No need to mess with DNS or deploy just to have others test out your changes.
Great for working with browser testing tools like browserling or external api callback services like twilio which require a public url for callbacks.


npm install -g localtunnel
This will install the localtunnel module globally and add the 'lt' client cli tool to your PATH.


Assuming your local server is running on port 8000, just use the lt command to start the tunnel.
lt --port 8000
Thats it! It will connect to the tunnel server, setup the tunnel, and tell you what url to use for your testing. This url will remain active for the duration of your session; so feel free to share it with others for happy fun time!
You can restart your local server all you want, lt is smart enough to detect this and reconnect once it is back.


Below are some common arguments. See lt --help for additional arguments
  • --subdomain request a named subdomain on the localtunnel server (default is random characters)
  • --local-host proxy to a hostname other than localhost


The localtunnel client is also usable through an API (for test integration, automation, etc)

localtunnel(port [,opts], fn)

Creates a new localtunnel to the specified local portfn will be called once you have been assigned a public localtunnel url. opts can be used to request a specific subdomain.
var localtunnel = require('localtunnel');

var tunnel = localtunnel(port, function(err, tunnel) {
    if (err) ...

    // the assigned public url for your tunnel
    // i.e.

tunnel.on('close', function() {
    // tunnels are closed


  • subdomain A string value requesting a specific subdomain on the proxy server. Note You may not actually receive this name depending on availablily.
  • local_host Proxy to this hostname instead of localhost. This will also cause the Host header to be re-written to this value in proxied requests.


The tunnel instance returned to your callback emits the following events
errorerrfires when an error happens on the tunnel
close fires when the tunnel has closed
The tunnel instance has the following methods
close close the tunnel

other clients

Clients in other languages


See localtunne/localtunnel-server for details on the server that powers local tunnel.

 server for


Build Status
localtunnel exposes your localhost to the world for easy testing and sharing! No need to mess with DNS or deploy just to have others test out your changes.
This repo is the server component. If you are just looking for the CLI localtunnel app, see (


The default localtunnel client connects to the server. You can, however, easily set up and run your own server. In order to run your own localtunnel server you must ensure that your server can meet the following requirements:
  • You can set up DNS entries for your domain.tld and *.domain.tld (or sub.domain.tld and *.sub.domain.tld).
  • The server can accept incoming TCP connections for any non-root TCP port (i.e. ports over 1000).
The above are important as the client will ask the server for a subdomain under a particular domain. The server will listen on any OS-assigned TCP port for client connections.


# pick a place where the files will live
git clone git://
cd localtunnel-server
npm install

# server set to run on port 1234
bin/server --port 1234
The localtunnel server is now running and waiting for client requests on port 1234. You will most likely want to set up a reverse proxy to listen on port 80 (or start localtunnel on port 80 directly).

use your server

You can now use your domain with the --host flag for the lt client.
lt --host http://sub.example.tld:1234 --port 9000
You will be assigned a URL similar to
If your server is acting as a reverse proxy (i.e. nginx) and is able to listen on port 80, then you do not need the :1234 part of the hostname for the lt client.


You can deploy your own localtunnel server using the prebuilt docker image.
Note This assumes that you have a proxy in front of the server to handle the http(s) requests and forward them to the localtunnel server on port 3000. You can use our localtunnel-nginx so accomplish this.
If you do not want ssl support for your own tunnel (not recommended), then you can just run the below with --port 80 instead.
docker run -d \
    --restart always \
    --name localtunnel \
    --net host \
    defunctzombie/localtunnel-server:latest --port 3000