Skip to content


Repository files navigation


Coverage Status PyPI version PyPI - Python Version License: MIT

This is the Python client for Unleash. It implements Client Specifications 1.0 and checks compliance based on spec in unleash/client-specifications

What it supports:

  • Default activation strategies using 32-bit Murmurhash3
  • Custom strategies
  • Full client lifecycle:
    • Client registers with Unleash server
    • Client periodically fetches feature toggles and stores to on-disk cache
    • Client periodically sends metrics to Unleash Server
  • Tested on Linux (Ubuntu), OSX, and Windows

Check out the project documentation and the changelog.


Check out the package on Pypi!

pip install UnleashClient



from UnleashClient import UnleashClient

client = UnleashClient(
    custom_headers={'Authorization': '<API token>'})


For more information about configuring UnleashClient, check out the project reference docs!

Checking if a feature is enabled

A check of a simple toggle:


To supply application context, use the second positional argument:

app_context = {"userId": ""}
client.is_enabled("user_id_toggle", app_context)

Fallback function and default values

You can specify a fallback function for cases where the client doesn't recognize the toggle by using the fallback_function keyword argument:

def custom_fallback(feature_name: str, context: dict) -> bool:
    return True

client.is_enabled("my_toggle", fallback_function=custom_fallback)

You can also use the fallback_function argument to replace the obsolete default_value keyword argument by using a lambda that ignores its inputs. Whatever the lambda returns will be used as the default value.

client.is_enabled("my_toggle", fallback_function=lambda feature_name, context: True)

The fallback function must accept the feature name and context as positional arguments in that order.

The client will evaluate the fallback function only if an exception occurs when calling the is_enabled() method. This happens when the client can't find the feature flag. The client may also throw other, general exceptions.

For more information about usage, see the Usage documentation.

Getting a variant

Checking for a variant:

context = {'userId': '2'}  # Context must have userId, sessionId, or remoteAddr.  If none are present, distribution will be random.

variant = client.get_variant("variant_toggle", context)

> {
>    "name": "variant1",
>    "payload": {
>        "type": "string",
>        "value": "val1"
>        },
>    "enabled": True
> }

For more information about variants, see the Variant documentation.


For development, you'll need to setup the environment to run the tests. This repository is using tox to run the test suite to test against multiple versions of Python. Running the tests is as simple as running this command in the makefile:

tox -e py311

This command will take care of downloading the client specifications and putting them in the correct place in the repository, and install all the dependencies you need.

However, there are some caveats to this method. There is no easy way to run a single test, and running the entire test suite can be slow.

Manual setup

First, make sure you have pip or pip3 installed.

Then setup your viritual environment:

Linux & Mac:

python3 -m venv venv
source venv/bin/activate

Windows + cmd:

python -m venv venv


python -m venv venv

Once you've done your setup, run:

pip install -r requirements.txt

Run the get-spec script to download the client specifications tests:


Now you can run the tests by running pytest in the root directory.

In order to run a single test, run the following command:


# example: pytest tests/unit_tests/


In order to lint all the files you can run the following command:

make fmt