Repository files navigation

A Python library for the Blink Camera system (Python 3.9+)

Like the library? Consider buying me a cup of coffee!

Buy me a Coffee!

BREAKING CHANGE WARNING:As of0.22.0 the library uses asyncio which will break any user scripts used prior to this version. Please see the updated examples below and theblinkapp.py orblinksync.py examples in theblinkapp/ directory for examples on how to migrate.

Disclaimer:Published under the MIT license - See LICENSE file for more details.

"Blink Wire-Free HS Home Monitoring & Alert Systems" is a trademark owned by Immedia Inc., seewww.blinkforhome.com for more information.I am in no way affiliated with Blink, nor Immedia Inc.

Original protocol hacking by MattTW :https://github.com/MattTW/BlinkMonitorProtocol

API calls faster than 60 seconds is not recommended as it can overwhelm Blink's servers. Please use this module responsibly.

pip install blinkpy

To install the current development version, perform the following steps. Note that the following will create a blinkpy directory in your home area:

$cd~$ git clone https://github.com/fronzbot/blinkpy.git$cd blinkpy$ pip install.

If you'd like to contribute to this library, please read thecontributing instructions.

This library was built with the intention of allowing easy communication with Blink camera systems, specifically to support theBlink component inhomeassistant.

The simplest way to use this package from a terminal is to callawait Blink.start() which will prompt for your Blink username and password and then log you in. In addition, http requests are throttled internally via use of theBlink.refresh_rate variable, which can be set at initialization and defaults to 30 seconds.

importasynciofromaiohttpimportClientSessionfromblinkpy.blinkpyimportBlinkasyncdefstart():blink=Blink(session=ClientSession())awaitblink.start()returnblinkblink=asyncio.run(start())

This flow will prompt you for your username and password. Once entered, if you likely will need to send a 2FA key to the blink servers (this pin is sent to your email address). When you receive this pin, enter at the prompt and the Blink library will proceed with setup.

In some cases, having an interactive command-line session is not desired. In this case, you will need to set theBlink.auth.no_prompt value toTrue. In addition, since you will not be prompted with a username and password, you must supply the login data to the blink authentication handler. This is best done by instantiating your own auth handler with a dictionary containing at least your username and password.

importasynciofromaiohttpimportClientSessionfromblinkpy.blinkpyimportBlinkfromblinkpy.authimportAuthasyncdefstart():blink=Blink(session=ClientSession())# Can set no_prompt when initializing auth handlerauth=Auth({"username":<yourusername>,"password":<yourpassword>},no_prompt=True)blink.auth=authawaitblink.start()returnblinkblink=asyncio.run(start())

Since you will not be prompted for any 2FA pin, you must call theblink.auth.send_auth_key function. There are two required parameters: theblink object as well as thekey you received from Blink for 2FA:

awaitauth.send_auth_key(blink,<yourkey>)awaitblink.setup_post_verify()

Other use cases may involved loading credentials from a file. This file must bejson formatted and contain a minimum ofusername andpassword. A built in function in theblinkpy.helpers.util module can aid in loading this file. Note, ifno_prompt is desired, a similar flow can be followed as above.

importasynciofromaiohttpimportClientSessionfromblinkpy.blinkpyimportBlinkfromblinkpy.authimportAuthfromblinkpy.helpers.utilimportjson_loadasyncdefstart():blink=Blink()auth=Auth(awaitjson_load("<File Location>"))blink.auth=authawaitblink.start()returnblinkblink=asyncio.run(start())

This library also allows you to save your credentials to use in future sessions. Saved information includes authentication tokens as well as unique ids which should allow for a more streamlined experience and limits the frequency of login requests. This data can be saved as follows (it can then be loaded by following the instructions above for supplying credentials from a file):

awaitblink.save("<File location>")

Cameras are instantiated as individualBlinkCamera classes within aBlinkSyncModule instance. All of your sync modules are stored within theBlink.sync dictionary and can be accessed using the name of the sync module as the key (this is the name of your sync module in the Blink App).

The below code will display cameras and their available attributes:

forname,camerainblink.cameras.items():print(name)# Name of the cameraprint(camera.attributes)# Print available attributes of camera

The most recent images and videos can be accessed as a bytes-object via internal variables. These can be updated with calls toBlink.refresh() but will only make a request if motion has been detected or other changes have been found. This can be overridden with theforce flag, but this should be used for debugging only since it overrides the internal request throttling.

camera=blink.cameras['SOME CAMERA NAME']awaitblink.refresh(force=True)# force a cache update USE WITH CAUTIONcamera.image_from_cache# bytes-like image object (jpg)camera.video_from_cache# bytes-like video object (mp4)

Theblinkpy api also allows for saving images and videos to a file and snapping a new picture from the camera remotely:

camera=blink.cameras['SOME CAMERA NAME']awaitcamera.snap_picture()# Take a new picture with the cameraawaitblink.refresh()# Get new information from serverawaitcamera.image_to_file('/local/path/for/image.jpg')awaitcamera.video_to_file('/local/path/for/video.mp4')

Methods exist to arm/disarm the sync module, as well as enable/disable motion detection for individual cameras. This is done as follows:

# Arm a sync moduleawaitblink.sync["SYNC MODULE NAME"].async_arm(True)# Disarm a sync moduleawaitblink.sync["SYNC MODULE NAME"].async_arm(False)# Print arm status of a sync module - a system refresh should be performed firstawaitblink.refresh()sync=blink.sync["SYNC MODULE NAME"]print(f"{sync.name} status:{sync.arm}")

Similar methods exist for individual cameras:

camera=blink.cameras["SOME CAMERA NAME"]# Enable motion detection on a cameraawaitcamera.async_arm(True)# Disable motion detection on a cameraawaitcamera.async_arm(False)# Print arm status of a sync module - a system refresh should be performed firstawaitblink.refresh()print(f"{camera.name} status:{camera.arm}")

You can also use this library to download all videos from the server. In order to do this, you must specify apath. You may also specifiy a how far back in time to go to retrieve videos via thesince= variable (a simple string such as"2017/09/21" is sufficient), as well as how many pages to traverse via thestop= variable. Note that by default, the library will search the first ten pages which is sufficient in most use cases. Additionally, you can specify one or more cameras via thecamera= property. This can be a single string indicating the name of the camera, or a list of camera names. By default, it is set to the string'all' to grab videos from all cameras. If you are downloading many items, setting thedelay parameter is advised in order to throttle sequential calls to the API. By default this is set to1 but can be any integer representing the number of seconds to delay between calls.

Example usage, which downloads all videos recorded since July 4th, 2018 at 9:34am to the/home/blink directory with a 2s delay between calls:

awaitblink.download_videos('/home/blink',since='2018/07/04 09:34',delay=2)

Since local storage is within a customer's residence, there are no guarantees for latencyand availability. As a result, the API seems to be built to deal with these conditions.

In general, the approach appears to be this: The Blink app has to query the syncmodule for all information regarding the stored clips. On a click to view a clip, the app asksfor the full list of stored clips, finds the clip in question, uploads the clip to thecloud, and then downloads the clip back from a cloud URL. Each interaction requires polling forthe response since networking conditions are uncertain. The app also caches recent clips and the manifest.

1. Request the local storage manifest be created by the sync module.
   • POST{base_url}/api/v1/accounts/{account_id}/networks/{network_id}/sync_modules/{sync_id}/local_storage/manifest/request
   • Returns an ID that is used to get the manifest.
2. Retrieve the local storage manifest.
   • GET{base_url}/api/v1/accounts/{account_id}/networks/{network_id}/sync_modules/{sync_id}/local_storage/manifest/request/{manifest_request_id}
   • Returns full manifest.
   • Extract the manifest ID from the response.
3. Find a clip ID in the clips list from the manifest to retrieve, and request an upload.
   • POST{base_url}/api/v1/accounts/{account_id}/networks/{network_id}/sync_modules/{sync_id}/local_storage/manifest/{manifest_id}/clip/request/{clip_id}
   • When the response is returned, the upload has finished.
4. Download the clip using the same clip ID.
   • GET{base_url}/api/v1/accounts/{account_id}/networks/{network_id}/sync_modules/{sync_id}/local_storage/manifest/{manifest_id}/clip/request/{clip_id}