Philips Hue lighting system allows you to control the lighting in your home from a mobile app, or from any application you write yourself that can connect to the internet. The center of the system is the Hue bridge, an Ethernet-connected device that communicates with compatible lamps through the ZigBee HA radio protocol. Philips makes a range of Hue-compatible lamps, and many third-party lamps work with the same protocol. IConnectHue maintains a thorough list of Hue-compatible lamps. Mostly Euro-market-centric, but still very useful.
Each Hue bridge is its own HTTP web server, and can be controlled using the Hue REST API. There are libraries to control the Hue available in many programming languages. The tutorials here are all in client-side JavaScript using P5.js, plain JavaScript, or server-side using node.js, or micrcontroller-based for Arduino.
The Hue API V2 was recently released, and will eventually supersede V1. These notes currently refer to V1 (19 Jan 2021) and still work. Hue API V2 features HTTPS, as well as some breaking changes from V1.
-
To get started programming Hue apps, you'll need access to a Hue bridge. You'll want a Hue account too. The developer accounts are free. Keep the Hue Developers Site link handy.
-
The Hue app for Android or iOS is helpful when developing, because it works when your project doesn't yet. The Hue Essentials app is a pretty helpful alternative also.
-
Every Hue has a debug interface, available at
http://$ADDR/debug/clip.htmlReplace$ADDRwith your bridge's IP address. When you're developing, you can use this to send API commands to the bridge to test things out. -
Peter Murray's node-hue-api library for node.js is the best of the various node.js JavaScript libraries I've tested for controlling the Hue.
-
For controlling the Hue from a browser client, p5.js does a good job, as it's got a simple http request API. You can use plain JavaScript as well, of course.
-
The ArduinoHTTPClient library and the Arduino_JSON library are useful if you're using any of the Arduino WiFi-enabled boards to connect to your Hue bridge. (Note: there's another JSON library by the same name with no underscore. That one is not the one used here).
-
The command line tool curl is helpful to test HTTP requests to your bridge, if you are familiar with the POSIX (linux/unix)command line. Curl's not available in the Windows command interface, but you can get it through the Windows Subsystem for Linux in Windows 10, or through cygwin, an application that provides a linux shell for Windows. Here are some details on using curl to access a hue bridge.
Other than these, you'll need to know some HTML and JavaScript, and a text editor, a command line interface, and a browser to try the examples on this site. The command line examples shown here are all for Linux or Unix, but will work on Windows 10 using the Windows Subsystem for Linux in Windows 10.
Your app will need to be identified to your bridge using a unique username. This example page will simplify that, if you know your bridge's IP address. You can't run this page from a remote server; download it and open the index.html file in a browser from your file system. This is because the Hue API version 1 uses HTTP, not HTTPS, and because your hub has a local IP address, and your web host does not .
You can also follow the Hue developer site's Getting Started instructions.
When you do it right, you should get a response like this:
[
{
"success": {
"username": "newusername"
}
}
]
Now you're ready to write code for your bridge. Regardless of what environment you're programming in, you'll use the username you established here.
The debug clip interface is a page that you can use to test functions on your bridge. Its address is:
http://your.bridge.ip.address/debug/clip.html
It can be used to send any API command to your bridge. The general query string is as follows:
/api/$HUE_USER/command
Generally, querying the state of the device is done using GET commands, and changing them is done using PUT. For example, to query the state of all your lights, enter the following in the URL field:
/api/$HUE_USER/lights
And click GET. To query the configuration, enter:
/api/$HUE_USER/config
And click GET. To turn on light 1, enter the following in the URL field:
/api/$HUE_USER/lights/1/state
And enter the following in the Message Body field:
{"on":true}
And click PUT. The light should come on, and the bridge should reply:
[
{
"success": {
"/lights/1/state/on": true
}
}
]
For more on the Hue API, see the Hue Getting Started guide, the Hue API Core Concepts, and the full Hue API documentation. You'll need to create a free account and log in to use the developer portal.
When you've added your bridge to your network, you should be able to use the Hue app or the Hue Essentials app to get the IP address. But on a complex network like a school network, that may not work. Your mobile device and your Hue bridge have to be on the same local network for this to work. For example, if your WiFi network is not the same local net as your wired Ethernet network (where the bridge lives), you may not be able to get the address. But if you can get the bridge's MAC address, then you can search for it on your network. Here's how
Every bridge has a unique You can find the MAC address on the bottom of your bridge. It's a six-byte number in hexadecimal notation like so:
00:17:88:0B:14:48
Some bridges will only show the last three bytes. For example, the bridge above might show just 0B1448. With older Hue bridges, the first three bytes will always be 00:17:88. With newer ones, you might also see EC:B5:FA instead.
To look for your bridge on your network, make sure you have the first three digits of the local network, and that you can access it, then open a command line interface and type:
$ ping -c 5 xxx.xxx.xxx.255
Where xxx.xxx.xxx are the first three numbers of your network. For example, on a network whose router is 172.16.130.1, you'd enter 172.16.130.255. Sometimes large institutions will use two different subnets for wired vs. wireless networks, but they will still be on the same larger local network.
You'll get a list of responses, as devices on the network respond to your ping requests. When it's done, type:
$ arp -a
You'll get a list of all the devices on the same network that your computer can see. Look for the one that matches the MAC address of your bridge, and the IP address next to it will be your bridge's IP address. The last three bytes of your MAC address from the label on the bottom. The first three will likely be either 00:17:88 or EC:B5:FA, as explained above. here's a typical example:
$ arp -a
? (192.168.0.1) at ac:b7:16:61:e3:77 on en0 ifscope [ethernet]
? (192.168.0.3) at 00:17:88:0B:14:48 on en0 ifscope [ethernet]
? (192.168.0.255) at ff:ff:ff:ff:ff:ff on en0 ifscope [ethernet]
In this case, the hue's IP address is 192.168.0.3.
You can search for new lamps on the bridge using the regular mobile Hue app. You can also do it from the debug clip interface using a POST request on the following URL:
/api/$HUE_USER/lights/
Fill in your bridge's address for $ADDR and your hue username for $HUE_USER. You should get a reply like this:
[[ { "success": { "/lights": "Searching for new devices" } }]
After 90 seconds, you can scan for new lamps that were added like so:
/api/$HUE_USER/lights/new
In curl:
$ curl -X GET http://$ADDR/api/$HUE_USER/lights/new
This will list only the new lamps added after a scan for new lamps.
If you're trying to add a lamp that was previously connected to a different bridge, you'll need to use a different approach. Place the lamp close to the bridge with which you want to control it (closer than any other lamp, ideally less than 1 meter). Turn off all other lamps connected to the bridge, or make sure they're much further away than the one you want.
Send the following the debug clip interface using a PUT request:
/api/$HUE_USER/config/
In the body of your request put:
{"touchlink": true}
In curl that's:
$ curl -X PUT -d '{"touchlink": true}' http://$ADDR/api/$HUE_USER/config
The lamp should blink a few times, and the server will respond with a success message. You can now add the lamp using the find new lamps request described above.
If touchlink doesn't succeed in resetting a lamp, you may need to reset it manually. It's different for each manufacturer, but the general pattern is to turn the lamp on and off three to five times, after which it will blink or dim slightly to indicate that it's been reset. For example, with the GE Link bulbs, the pattern is:
- Turn on for 3 sec.
- Turn off for 3 sec.
- Repeat steps 1 and 2 four more times
- Once the lamp blinks, it's factory reset and you can add it to a bridge normally.
For some third party bulbs, like the Cree Connect or the Innr bulbs, you can do it with 2 second intervals. When in doubt, go with three seconds, though.
If you have one of the Hue remote controls, you can reset a lamp using it as well. To do this, turn on the lamp and place the remote control within six inches of the lamp. Press the power on and power off buttons (or power on and scene select buttons on more recent models) similtaneously and hold them until the bulb flashes. This will put the lamp in its default factory state, ready to pair with a new bridge.
You can delete a lamp from a bridge from the debug clip interface using a DELETE request on the following URL:
/api/$HUE_USER/lights/1
In curl:
$ curl -X DELETE http://$ADDR/api/$HUE_USER/lights/1
Replace 1 with the number of the light you wish to delete.
To get the status of all connected lights, send the following from the debug clip interface using a GET request:
/api/$HUE_USER/lights/
In curl, that's:
$ curl http://$ADDR/api/$HUE_USER/lights/
This should return the list of available lights, like so:
{"1":{"state":{"on":true,"bri":254,"hue":14314,"sat":172,"effect":"none","xy":[0.4791,0.4139],"ct":405,"alert":"none","colormode":"ct","reachable":true},"type":"Extended color light","name":"Hue color light 1","modelid":"LCT001","manufacturername":"Philips","uniqueid":"00:17:88:01:00:ff:9a:28-0b","swversion":"5.127.1.26581"}}
To turn a light on you need to know which number it is. Then you change its state like so:
Send the following from the debug clip interface using a PUT request:
/api/$HUE_USER/lights/4/state
In the body of your request put:
{"on": true}
In curl, that's:
$ curl -X PUT -d '{"on":true}' http://$ADDR/api/$HUE_USER/lights/4/state
To turn it off, change the body of the request to
{"on": false}
You can change any of the properties of a light's state this way. Take a look, for example, at light 1 from the Getting the State of All Lights section above:
{"1":{"state":{"on":true,"bri":254,"hue":14314,"sat":172,"effect":"none","xy":[0.4791,0.4139],"ct":405,"alert":"none","colormode":"ct","reachable":true},
As long as the reachable property is true, meaning that the bridge tried to reach the lamp and got a response, you can change any of the other properties. This is a color lamp, and has three modes, hs (for hue, saturation), ct (for color temp), and xy (for x and y dimensions in the CIE1931 color space). If you change either the hue or saturation, the lamp's colormode changes to hs, and if you change the color temperature, the colormode changes to ct. If you send xy values, the colormode changes to xy mode.
- bri (brightness): 0-254
- hue: 0-65535, through the colorwheel from red to red
- sat (saturation): 0-254
- effect: "none" or "colorloop" see Developer API docs
- xy (x and y position in CIE1031 colorspace): 2 floats, 0.0 to 1.0 each
- ct (color temperature): in mired, 153 - 500
- alert - see Developer API docs
Different lights will have different properties in their state variable that you can change. The p5.js sketch will scan all the properties of each lamp on the
The MKR1000, MKR1010, and Nano 33 IoT Arduino models can control the Hue bridge via HTTP requests as well. There are some Arduino Hue examples, with notes, in this repository as well.
![@dependabot[bot]](https://avatars.githubusercontent.com/in/29110?s=64&v=4){kind=small}