How to Connect to Tableau Server Using Python

Instead of manually publishing data sources or managing users one by one, you can use Python to automate almost any Tableau Server task you can think of. Python scripts can refresh extracts on custom schedules, add or remove users in bulk, and programmatically publish content, saving your team countless hours. This tutorial will walk you through exactly how to connect to your Tableau Server using Python, authenticate your credentials, and start interacting with your projects, workbooks, and data sources.

Graphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Build Your First Agent

Why Bother Connecting to Tableau Server with Python?

Connecting to Tableau with Python isn’t just about showing off your coding skills, it's a powerful way to make your data and analytics workflow more efficient, scalable, and integrated. If you find yourself doing the same tasks over and over in the Tableau UI, chances are you can automate them.

Here are just a few benefits:

• Automate Repetitive Tasks: Stop the manual grind. You can write scripts to publish dozens of workbooks, refresh data extracts based on triggers from other systems, or update user permissions programmatically. Imagine an extract refreshing automatically the moment an ETL job finishes, not at a fixed time.
• Integrate Tableau into Data Pipelines: Python is the lingua franca of data engineering. By connecting it to Tableau, you can make Tableau a seamless part of a larger data pipeline. For example, a Python script that scrapes web data, cleans it in pandas, and runs a machine learning model can push the final, processed data directly to a Tableau data source.
• Dynamic Content Creation: You can programmatically download workbooks, modify their XML to change data source connections or apply new formatting, and then republish them. This allows you to scale content for different regions, departments, or clients without manual duplication.
• Advanced Server Management: Need a special report of all data sources that haven't been recently used? Want to enforce specific tagging conventions across all your content? Python makes it possible to perform complex administrative tasks that aren't available through the standard UI.

The Easiest Method: Using the tableau-server-client Library

While you could technically interact with Tableau Server by making direct requests to its REST API using a library like requests, there's a much easier way. Tableau officially maintains a Python library called tableau-server-client that handles all the complicated API calls for you.

This library acts as a user-friendly wrapper. Instead of you figuring out the exact API endpoints, authentication headers, and request formats, the library provides simple Python objects and methods like server.projects.get(). It simplifies development, reduces bugs, and is actively maintained to keep up with new Tableau Server features.

In this guide, we'll focus exclusively on using this library, as it’s the recommended approach for virtually all use cases.

Getting Started: Prerequisites & Setup

Before you can write a single line of Python, you need a few things in place. Let’s get your environment set up properly.

1. Install the Python Library

First, you need to install the tableau-server-client library. You can do this easily using pip from your terminal or command prompt.

pip install tableau-server-client

Running this command fetches the latest version of the library from the Python Package Index (PyPI) and installs it on your machine, along with its dependencies.

AI Agents for Marketing

Crash Course

Graphed

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Download the PDF

2. Obtain Your Tableau Server Credentials

To connect to the server, you’ll need to prove you have access. We highly recommend using a Personal Access Token (PAT) for automated tasks instead of your username and password. PATs are more secure, as they can be revoked at any time without impacting your regular login, and they don't expose your primary credentials in scripts.

You’ll need the following four pieces of information:

• Server URL: The main address you use to access Tableau in your browser (e.g., https://tableau.mycompany.com).
• Site ID (Content URL): When you log into Tableau Server, the site-specific identifier appears in the browser's URL right after /site/. For example, if your URL is https://tableau.mycompany.com/#/site/MarketingAnalytics/..., your site ID is MarketingAnalytics. If you are using the "Default" site (the one that comes with every server), your site ID should be an empty string ('').
• Personal Access Token (PAT) Name: When you create a PAT in your Tableau User Settings, you give it a name to help you remember its purpose (e.g., python_automation_script).
• PAT Secret: This is the actual token. Tableau only shows this to you once upon creation. Make sure you copy and store it in a secure location, like an environment variable or a secrets management tool.

With the library installed and your credentials handy, you're ready to start building your connection script.

A Step-by-Step Guide to Connecting with Python

Connecting to the server involves three main steps: creating a server object, building an authentication object, and then signing in. We'll break it down piece by piece.

Step 1: Import the Necessary Libraries

First, start your Python script by importing the tableau_server_client library. It's common practice to give it an alias, such as TSC, to make your code shorter and more readable.

import tableau_server_client as TSC

Step 2: Define Your Server and Authentication Details

Next, define variables to hold your server URL and PAT credentials. Putting them at the top of your script makes them easy to find and change later. Never hardcode sensitive credentials directly if your code will be shared, use environment variables instead.

SERVER_URL = 'https://YOUR_TABLEAU_SERVER_URL.com' 
PAT_NAME = 'your_personal_access_token_name'
PAT_SECRET = 'your_super_secret_personal_access_token'
SITE_ID = 'your_site_id'  # Use an empty string '' for the 'Default' site

Remember to replace the placeholder values with your actual information. If you're on the default server site, make sure SITE_ID is ''.

Step 3: Create the Tableau Server Object

Now, create an instance of the Server object from the library. This object represents your Tableau Server and is what you'll use to make all future requests.

# Create a server object
tableau_server = TSC.Server(SERVER_URL, use_server_version=True)

Setting use_server_version=True is a helpful feature. It tells the library to automatically detect the API version of your server and use the correct supported features, preventing compatibility issues.

Step 4: Authenticate with Your Personal Access Token

Here, you'll create the PersonalAccessTokenAuth object using your PAT details. Then, you'll sign in using a Python with statement. The with statement is the best way to handle signing in because it automatically signs you out when your code block is finished, which is great for managing server sessions.

# Create an authentication object
personal_access_token = TSC.PersonalAccessTokenAuth(
    token_name=PAT_NAME,
    personal_access_token=PAT_SECRET,
    site_id=SITE_ID
)

# The 'with' statement will sign you in and sign you out when you are done
with tableau_server.auth.sign_in(personal_access_token):
    # This block of code is executed while you are signed into the server
    print(f"Successfully connected to site: {tableau_server.site_id}")
    print(f"Server version: {tableau_server.version}")

Graphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Build Your First Agent

Putting It All Together: A Complete Connection Script

Here is the full, clean script that you can use as a template. Just copy, paste, and update the credential variables to test your connection.

import tableau_server_client as TSC

# --- Step 1: Define Connection Credentials ---
SERVER_URL = 'https://YOUR_TABLEAU_SERVER_URL.com' 
PAT_NAME = 'your_personal_access_token_name'
PAT_SECRET = 'your_super_secret_personal_access_token'
SITE_ID = 'your_site_id'  # Use an empty string '' for the Default site

# --- Step 2: Set up Server and Authentication ---
# Create a server object
tableau_server = TSC.Server(SERVER_URL, use_server_version=True)

# Create an authentication object using a Personal Access Token
personal_access_token = TSC.PersonalAccessTokenAuth(
    token_name=PAT_NAME,
    personal_access_token=PAT_SECRET,
    site_id=SITE_ID
)

# --- Step 3: Connect to Server ---
# The 'with' statement will sign you in and out automatically
with tableau_server.auth.sign_in(personal_access_token):
    print('Connection successful!')
    print(f"Server name: {tableau_server.server_info.product_version}")
    
    # Do some work with the server inside this block
    all_projects, pagination_item = tableau_server.projects.get()
    print("\nProjects on this site:")
    for project in all_projects:
        print(f" - {project.name} (ID: {project.id})")

If the script runs without errors and prints your list of projects, congratulations - you have successfully connected to Tableau Server with Python!

What Next? Common Automation Tasks

Once you are connected, the possibilities are endless. The tableau_server object is your gateway to managing nearly everything on the server. Here are a few simple examples.

Finding a Specific Workbook

Often, you'll need the ID of a specific asset before you can interact with it. Here’s how you could find a workbook named “Quarterly Sales Dashboard.”

workbook_name_to_find = 'Quarterly Sales Dashboard'

with tableau_server.auth.sign_in(personal_access_token):
    all_workbooks, pagination = tableau_server.workbooks.get()

    target_workbook = None
    for wb in all_workbooks:
        if wb.name == workbook_name_to_find:
            target_workbook = wb
            break
            
    if target_workbook:
        print(f"Found workbook '{target_workbook.name}' with ID: {target_workbook.id}")
    else: 
        print(f"Workbook '{workbook_name_to_find}' not found.")

AI Agents for Marketing

Crash Course

Graphed

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Download the PDF

Publishing a Hyper File

Let’s say an automated Python script generates a .hyper file with fresh data. You can publish it directly to Tableau Server. You'll first need the project_id you want to publish it to (which you can get with the method shown earlier).

# You must know the project ID first
target_project_id = 'PROJECT_ID_FROM_PREVIOUS_EXAMPLE'
hyper_file_path = '/path/to/your/data.hyper'

with tableau_server.auth.sign_in(personal_access_token):
    # Create a DatasourceItem to define its name and project destination
    new_datasource = TSC.DatasourceItem(name='My Automated Data', project_id=target_project_id)
    
    # Publish the hyper file, overwriting if it already exists
    published_datasource = tableau_server.datasources.publish(
        new_datasource,
        hyper_file_path,
        TSC.Server.PublishMode.Overwrite
    )
    print(f"Data source '{published_datasource.name}' published successfully.")

Common Pitfalls and Troubleshooting Tips

If you run into issues, don't worry. Here are solutions to the most common problems:

• Connection Errors: If you get a requests.exceptions.ConnectionError, the most common cause is either an incorrect server URL or a network firewall blocking your Python script from reaching the server. Check if you can access the URL in your web browser from the same machine.
• Authentication Errors (401 Unauthorized): This error means the server rejected your credentials. Double-check your PAT name, PAT secret, and especially the site_id. Remember, for the 'Default' site, the site_id is an empty string, not the word "Default."
• SSL Certificate Errors: Many corporate Tableau Servers use internal SSL certificates. You may see a requests.exceptions.SSLError. For a quick fix during testing (not recommended in production), you can disable SSL verification. Just note this is less secure:

tableau_server.add_http_options({'verify': False})
# Sign in directly *after* setting this option
with tableau_server.auth.sign_in(personal_access_token):
    ...

• Site Not Found Errors: If you get a message saying the site could not be found, your site_id is almost certainly incorrect. Check the URL in your browser again to find the correct value.

Final Thoughts

Connecting to your Tableau Server with Python bridges the gap between manual analytics and true automation. Using the tableau-server-client library, you can programmatically manage your server, streamline data publishing, and deeply integrate Tableau into your data workflows, saving you hours of repetitive clicking and manual oversight.

Of course, staying on top of your analytics shouldn't always require learning new libraries or writing custom scripts. At Graphed, we harness AI so you don't even have to write code. By simply asking questions in ordinary language, you connect your key data sources — from Salesforce to Google Ads — and watch as real-time dashboards and reports are created for you in moments. It’s an accessible way to centralize your key metrics and unlock insights without ever leaving the conversation.