An unofficial, fully functional Python SDK for the Unipile API.

Integrating with powerful APIs can be complex, unfortunately Unipile does not provide an official Python client (actual for 2025), only Node.JS based official client.

When I started working with the Unipile API and Python to automate LinkedIn interactions, I found myself writing a lot of boilerplate code to handle authentication, make HTTP requests, and parse responses. It was repetitive and error-prone. I wanted a simpler way to interact with the API, so I could focus on building features, not on the underlying HTTP calls.

That's why I built this SDK. It's a Pythonic wrapper around the Unipile API that handles the heavy lifting for you. Now, you can retrieve user profiles, search for candidates, and send messages with just a few lines of code. It's the tool I wish I had when I started, and I hope it makes your life easier too.

Get up and running with the Unipile Python SDK in just a few steps.

Demonstration of the work:

pip install i-unipile-sdk  # or add it into requirements and install it

Alternatively, you can use uv and add Unipile SDK into your existing project.

uv add i-unipile-sdk

You need to get some data before processing:

• Get DNS (base_url), can be copied from dashboard.
• Generate your API token (auth) in access tokens
• Connect account and get its id (default_account_id), can be copied next to account name in the accounts section.

All related information can be found in the getting-started documentation. I recommended watching there a quick start video first.

You can configure the client by passing an ClientOptions object directly:

from unipile_sdk import Client
from unipile_sdk.client import ClientOptions

options = ClientOptions(
    auth="your_api_key_here",  # Your Unipile API Key
    base_url="https://your_base_url_here",  # The Unipile API base URL
    default_account_id="your_account_id_here",  # The ID of the account to use
)

client = Client(options=options)

me = client.users.me()
print(f"My occupation is: {me.occupation}")

This section showcases some of the more advanced features of the SDK. For brevity, it is assumed that the client object has been initialized as shown in the Quick Start section.

To use examples, you need to import the relevant models and functions from the SDK, e.g., LinkedinSearchPayload, iterate_paginated_api, and Client:

from unipile_sdk.models import LinkedinSearchPayload, NotFoundType
from unipile_sdk.helpers import iterate_paginated_api
from unipile_sdk.errors import APIResponseError

Retrieve a list of all connected accounts.

accounts = client.accounts.accounts(limit=100)
for account in accounts.items:
    print(f"Account: {account.type} - {account.name}")

Perform a detailed search for people on LinkedIn. The SDK will handle pagination for you.

payload = LinkedinSearchPayload(
    api="classic", category="people", keywords="Software Engineer, Programmer"
)

for person in iterate_paginated_api(
    client.ln_search.search, payload=payload, max_total=10
):
    print(f"Found Person: {person.name} ({person.id})")

Fetch company details using its name.

company = client.ln_search.retrieve_company(identifier="LinkedIn")
print(f"Company: {company.name} ({company.id})")

For more examples, please check the tests/integration directory. The following examples are also available as Python files in the examples directory.

This SDK is currently under development and should be considered a work in progress. Here are a few things to keep in mind:

• Incomplete API Coverage: The SDK does not cover the entire Unipile API. The primary focus has been on the LinkedIn-related endpoints.
• Potential for Bugs: As with any active development project, there may be bugs or incomplete features.

Contributions to improve the SDK and expand its coverage are welcome.

Contributions are welcome! Here's how you can get started with the development of the SDK.

1. Clone the repository.

2. Install the development dependencies:

   git clone repo-url
   cd repo-directory
   uv pip install -e ".[dev]" # or `pip install -e ".[dev]"`

To run the tests, you need to set up your test environment variables. Create a .tests.env file in the root of the project:

UNIPILE_BASE_URL should be without protocol (like api.unipile.com...).

UNIPILE_ACCESS_TOKEN=your_api_token_here
UNIPILE_BASE_URL=your_unipile_api_basi_url
UNIPILE_ACCOUNT=your__account_id_here

UNIPILE_LN_DEFAULT_USER_URN=your_default_user_urn_here
UNIPILE_LN_USER_URN_TO_MESSAGE=your_target_user_urn_here

Then, run the tests using pytest:

pytest

Some tests that perform actions like sending messages are marked as messaging and are skipped by default. To run them, use the --activities flag:

If you specify UNIPILE_LN_USER_URN_TO_MESSAGE, you can test messaging functionality by sending a message to that user URN. Be careful this can actually send a many same messages on LinkedIn. Use a test account or a controlled recipient.

There additional test_send_message_to_attendees test, which sends messages with specific logic. If there found only one message (usually invite message), it sends a test message to that attendee and asserts success. So expect some "thank you for connection" messages if --activities flag is used.

pytest --messaging