Lightweight, super fast (uses C and Rust libraries) pythonic wrapper for Betfair API-NG allowing all betting operations (including market and order streaming) and account operations, see examples.

docs

join slack group

Currently tested on Python 3.6, 3.7, 3.8 and 3.9.

$ pip install betfairlightweight

To use C/Rust libraries install with

$ pip install betfairlightweight[speed]

In order to connect to the Betfair API you will need an App Key, SSL Certificates and a username/password.

Follow these instructions to get your app key, you can either use a delayed or live key.

Follow these instructions to set up your SSL certificates. Save your .ctr and .key files to a local directory. The default directory where the library is looking for the keys is '/certs' but you can specify any other directory.

The library can then be used as follows:

import betfairlightweight

trading = betfairlightweight.APIClient('username', 'password', app_key='app_key', certs='/certs')

trading.login()

or the following for interactive login with no certs (not as secure)

import betfairlightweight

trading = betfairlightweight.APIClient('username', 'password', app_key='app_key')

trading.login_interactive()

event_types = trading.betting.list_event_types()

[<EventTypeResult>, <EventTypeResult>, ..]

Following endpoints are available:

• trading.login

• trading.login_interactive

• trading.keep_alive

• trading.logout

• trading.betting

• trading.account

• trading.navigation

• trading.scores

• trading.streaming

• trading.historical

• trading.in_play_service

• trading.race_card

Currently two listeners available, below will run the base listener which prints anything it receives. Stream listener is able to hold an order stream or a market stream (one per listener). The listener can hold a cache and push market_books/order_books out via a queue.

Exchange Stream API

from betfairlightweight.filters import (
    streaming_market_filter,
    streaming_market_data_filter,
)

betfair_socket = trading.streaming.create_stream()

market_filter = streaming_market_filter(
    event_type_ids=['7'],
    country_codes=['IE'],
    market_types=['WIN'],
)
market_data_filter = streaming_market_data_filter(
    fields=['EX_ALL_OFFERS', 'EX_MARKET_DEF'],
    ladder_levels=3
)

betfair_socket.subscribe_to_markets(
    market_filter=market_filter,
    market_data_filter=market_data_filter,
)

betfair_socket.start()  # blocking

The historic endpoint provides some basic abstraction for the historicdata api:

Historic Data API

trading.historic.get_my_data()

[{'plan': 'Basic Plan', 'purchaseItemId': 1343, 'sport': 'Cricket', 'forDate': '2017-06-01T00:00:00'}]

Taking advantage of the streaming code lightweight can parse/output historical data in the same way it process streaming data allowing backtesting or with a custom listener, csv creation (see examples).

Historic Data

stream = trading.streaming.create_historical_stream(
    file_path='horse-racing-pro-sample',
)

stream.start()

or use the stream generator:

stream = trading.streaming.create_historical_generator_stream(
    file_path='horse-racing-pro-sample',
)

g = stream.get_generator()

for market_books in g():
    print(market_books)