Moved (and renamed) to https://github.com/tesladdicts/testatus
Simple Python class to access the Tesla JSON API.
Written by Greg Glockner
This is a simple Python interface to the Tesla JSON API. With this, you can query your vehicle, control charge settings, turn on the air conditioning, and more. You can also embed this into other programs to automate these controls.
The class is designed to be simple. You initialize a Connection object, retrieve the list of Vehicle objects, then perform get/set methods on a Vehicle. There is a single get method [Vehicle.data_request()] and a single set method [Vehicle.command()] so that the class does not require changes when there are minor updates to the underlying JSON API.
This has been tested with Python 2.7 and Python 3.5. It has no dependencies beyond the standard Python libraries.
- Download the repository zip file and uncompress it
- Run the following command with your Python interpreter:
python setup.py install
Alternately, add the teslajson.py code to your program.
Connection(email, password, **kwargs):
Initialize the connection to the Tesla Motors website.
Required parameters:
-
Option one:
- email: your login for teslamotors.com
- password: your password for teslamotors.com
-
Option two: (May be combined with option one)
- tokens_file: A file containing json token authentication data as tesla generates. Updated when expires.
-
Option three:
- access_token: An active access token for your account. May expire.
Optional parameters:
- proxy_url: URL for proxy server
- proxy_user: username for proxy server
- proxy_password: password for proxy server
- retries: number of times to retry request before failing
- retry_delay: multiplicative backoff on failure
- tesla_client: Override API retrevial from pastebin
- debug: Activate HTTP debugging
Connection.vehicles: A list of Vehicle objects, corresponding to the
vehicles associated with your account on teslamotors.com.
Vehicle: The vehicle class is a subclass of a Python dictionary
(dict). A Vehicle object contains fields that identify your
vehicle, such as the Vehicle Identification Number (Vehicle['vin']).
All standard dictionary methods are supported.
Vehicle.wake_up(): Wake the vehicle.
Vehicle.data_all(): Retrieve all data values associated with vehicle.
Vehicle.data_request(name): Retrieve data values specified by name, such
as charge_state, climate_state, vehicle_state. Returns a
dictionary (dict). For a full list of name values, see the GET
commands in the Tesla JSON API.
Vehicle.command(name): Execute the command specified by name, such
as charge_port_door_open, charge_max_range. Returns a
dictionary (dict). For a full list of name values, see the POST commands
in the Tesla JSON API.
import teslajson
c = teslajson.Connection('youremail', 'yourpassword')
v = c.vehicles[0]
v.wake_up()
v.data_request('charge_state')
v.command('charge_start')
c = teslajson.Connection(access_token='b5bb9d8014a0f9b1d61e21e796d78dccdf1352f23cd32812f4850b878ae4944c', tesla_client='{"v1": {"id": "e4a9949fcfa04068f59abb5a658f2bac0a3428e4652315490b659d5ab3f35a9e", "secret": "c75f14bbadc8bee3a7594412c31416f8300256d7668ea7e6e7f06727bfb9d220", "baseurl": "https://owner-api.teslamotors.com", "api": "/api/1/"}}')
./teslajson.py --email my@email.com --pasword thepass --tokens_file /tmp/tesla.creds get
./teslajson.py --tokens_file /tmp/tesla.creds --vid 0 get
./teslajson.py --tokens_file /tmp/tesla.creds --retries 10 do wake_up
./teslajson.py --tokens_file /tmp/tesla.creds get climate_state
./teslajson.py --tokens_file /tmp/tesla.creds get gui_settings
./teslajson.py --tokens_file /tmp/tesla.creds get mobile_enabled
./teslajson.py --tokens_file /tmp/tesla.creds get data
./teslajson.py --tokens_file /tmp/tesla.creds do charge_port_door_open
Many thanks to Tim Dorr for documenting the Tesla JSON API. This would not be possible without his work.
This software is provided as-is. This software is not supported by or endorsed by Tesla Motors. Tesla Motors does not publicly support the underlying JSON API, so this software may stop working at any time. The author makes no guarantee to release an updated version to fix any incompatibilities.
Written by Seth Robertson
Use the teslajson library to do smart polling of your Tesla(s) and log the resulting JSON information to a directory for post-processing. Monitor different things with different frequency depending on what you are doing (e.g. driving, charging, pre-heating, nothing, etc).
Also support an insecure network API (so only allow local connections
to use it!) with --command 127.0.0.1:60001 (or some other local
address:port combination to bind to) to allow specific commands to be
executed on your behalf. Currently only "quit" and "autocondition"
are supported. Autocondition requests additional change and turns on
car climate control, to (e.g.) charge/heat the battery and get the
interior heat/cooled as needed, soon before leaving. After leaving, it
resets the desired battery charge back to normal.
Command line arguments are requried for authentication: --token,
--tokenfile, or --email and --password.
In order to log the data, supply an output directory with --outdir path. In addition to a file named by YEAR-MON-DAY.json, there is a
symlink cur.json to the most recent file.
You may override the intervals of important (polling frequency mostly)
by using --intervals inactive=61 or similar.
tesla-parser.py was created to read the stored data.
Example usage: tesla-parser.py /path/to/cur.json
By default it provides summary information for drives you make, charges you do, and standby times.
If you want to see more detailed information, use -v. Add another
-v or two to learn about even less important activity.
You may supply multiple files, and you may use the -f argument to
follow a file as it is appended to. A convenient way to dump all
historical information and then start printing any future information
is:
tesla-parser.py -f /var/logs/tesla/cur.json -n 0 /var/logs/tesla/20*.json
Example output:
2018-07-07 08:50:56 +0:20:04 Drove 20.58M at cost of 10% 25.4M at 80.9% efficiency
2018-07-07 09:16:00 +4:55:03 Sat&Lost 2% 15.8M or 77.2M/d
2018-07-07 14:11:03 +0:22:08 Drove 20.33M at cost of 8% 20.2M at 100.4% efficiency
2018-07-07 14:38:11 +0:12:12 Sat&Lost 0% 0.0M or 0.0M/d
2018-07-07 14:50:23 +2:32:50 Charged 25% (to 88%) 19.37kW 79.1M ( 31mph, 77.5kW 311.8M max)
In the above example, the vehicle started on a 20 mile trip at 8:50 am, which took 20 minutes. It actually use 25 rated miles of range (10% of battery), meaning 81% efficiency.
The vehicle then sat for 5 hours and lost 2% of charge (16 rated miles) for a 77 miles per day effective rate (perhaps cabin temperature control was enabled).
There was a 22 minute return trip, this time at 100% efficiency. The vehicle sat for 12 minutes (not losing any power), and then was charged for 2.5 hours, to 88%, getting 19kW or 79 rated miles. The charge was obtained at an average speed of 31mph, and implied a fully charged battery size of 78kW and a 312 maximum rated mile range.
tesla-parser.py is able to insert the stored data into a relational
database which may be more convenient for analysis. In order to do this
an adequate Postgresql database needs to be installed and accessible.
Create a database named "tesladata" and set a user (eg. "teslauser")
with privileges to create tables. Then run the file create_tables
with this user. Set the connection details in the file dbconfig,
including the password for the user (it is advisible to have that file
protected from other users so as not to reveal the password).
create user teslauser with encrypted password 'example';
create database tesladata;
grant all privileges on database tesladata to teslauser;
psql -U testauser testladata < create_tables.sql
To store the data into the database run tesla-parser.py with the
command line option --dbconfig dbconfig. Instead of dumping
summary statistics this will instead insert the data into
the database. Make sure to indicate the file(s) with the data desired.
For example:
tesla-parser.py -n 0 --dbconfig dbconfig /var/logs/tesla/20*.json
If the data had already been inserted into the database in a previous run, the program will issue appropriate warnings.
The poller_rpc.py program implements a client side of the RPC. It
is pretty primitive. Specify the remote address of the server with
--cmd_address 127.0.0.1:60001 or similar matching addressing. Then
you use --variables to specify commands. Example:
# Ask server to quit
./poller_rpc.py --cmd_address 127.0.0.1:60001 --variables cmd=quit
# Do default autoconditioning
./poller_rpc.py --cmd_address 127.0.0.1:60001 --variables cmd=autocondition
# Set battery charge limit to 77%, do NOT do cabin preconditioning
./poller_rpc.py --cmd_address 127.0.0.1:60001 --variables cmd=autocondition --variables level=77 --variables temp=
# Set battery charge limit to 90%, reset limit to 77% when next driving, precondition cabin to 80°
./poller_rpc.py --cmd_address 127.0.0.1:60001 --variables cmd=autocondition --variables level=90 --variables autoresetlimit=77 --variables temp=80
Only tested with one vehicle.