CORE Getting Started: Account, Keys, and First Request (2026)

Getting started overview

To begin using CORE, which encompasses the Wolfram Language and its various deployment options, you generally need to establish a Wolfram ID, obtain appropriate licensing or access credentials, and then interact with the Wolfram Engine. This interaction can be through a local installation of Mathematica, a cloud-based service, or via an API. The Wolfram Language is designed for symbolic computation, numerical analysis, and data visualization, offering a unified ecosystem for technical computing Wolfram Language reference documentation.

This guide focuses on the streamlined process for developers, particularly those looking to integrate Wolfram Language capabilities into external applications or scripts. This involves either setting up the Wolfram Engine for Developers locally or accessing cloud-based computation resources. The Wolfram Engine for Developers provides a free, limited-use license that is suitable for prototyping and non-commercial development Wolfram pricing and licensing options.

Here's a quick reference table to guide you through the initial setup process:

Create an account and get keys

The first step to interacting with CORE's computational capabilities is to create a Wolfram ID. This ID serves as your universal login for all Wolfram services, including access to the Wolfram User Portal, software downloads, and cloud resources.

1. Register for a Wolfram ID: Navigate to the Wolfram website and look for a "Create Account" or "Sign In" option. You will typically be prompted to provide an email address and create a password.
2. Verify Email: After registration, Wolfram Research will send a verification email to the address you provided. Follow the instructions in the email to activate your Wolfram ID.
3. Obtain Wolfram Engine for Developers License: For most development purposes, especially prototyping or non-commercial use, the Wolfram Engine for Developers is the recommended starting point. Once logged in with your Wolfram ID, you can request this free license. This usually involves completing a short form and agreeing to the terms of use. The license allows you to use the Wolfram Engine on a local machine without a full Mathematica front end.
4. Download and Install the Wolfram Engine: After securing your license, you will be directed to a download page within the Wolfram User Portal. Select the correct version for your operating system (Windows, macOS, Linux) and proceed with the installation. The installer will guide you through the process.
5. Activate the Engine: Upon the first execution of the Wolfram Engine (e.g., by running wolfram in your terminal), you will be prompted to activate it using your Wolfram ID credentials. This links your local installation to your Wolfram account and validates your license.
6. Cloud Access (Optional but Recommended for API): If you plan to use the Wolfram Cloud for API access, your Wolfram ID also grants you access to cloud notebooks and deployment capabilities. Cloud APIs often don't require explicit "keys" in the traditional sense like REST APIs, but rather rely on authenticated sessions or deployed functions accessible via unique URLs. For programmatic access to deployed cloud functions, you might generate an API key or use authentication tokens if your deployment requires it. Consult the Wolfram Cloud Deployment Guide for specifics on securing cloud-deployed APIs.

Your first request

Once the Wolfram Engine is installed and activated, you can execute your first Wolfram Language command. This can be done directly from your terminal, or by using one of the available SDKs, such as the Wolfram Client for Python.

Using the Command Line (Local Engine)

The simplest way to verify your installation is to run a command directly in your terminal:

1. Open your terminal or command prompt.
2. Type wolfram and press Enter. This will launch the Wolfram Language kernel.
3. At the In[1]:= prompt, type a simple command, for example, 2 + 2, and press Shift+Enter (or Enter if your terminal is configured for line-by-line input).
4. You should see the output Out[1]= 4.
5. To exit the kernel, type Exit[] and press Enter.

You can also execute a single command non-interactively:

wolfram -run "Print[FactorInteger[123456789]]"

This command will print the prime factorization of 123456789 to your console.

Using Wolfram Client for Python

For programmatic integration, the Wolfram Client for Python is a robust option. First, ensure you have Python installed, then install the client:

pip install wolframclient

Next, create a Python script (e.g., first_request.py):

from wolframclient.evaluation import WolframLanguageSession
from wolframclient.language import wl, wlexpr

# Initialize a session with the local Wolfram Engine
# Ensure the Wolfram Engine is installed and activated
session = WolframLanguageSession()

try:
    # Evaluate a simple expression
    result = session.evaluate(wl.FactorInteger(123456789))
    print(f"Factorization: {result}")

    # Evaluate a more complex expression using wlexpr
    complex_result = session.evaluate(wlexpr('Integrate[Sin[x]^2, x]'))
    print(f"Integral: {complex_result}")

finally:
    # Always terminate the session
    session.terminate()

Run the script from your terminal:

python first_request.py

You should see output similar to:

Factorization: {{3, 2}, {41, 1}, {271, 1}, {9091, 1}}
Integral: 1/2 (x - Cos[x] Sin[x])

This demonstrates successful communication with the Wolfram Engine from a Python script, executing both numerical and symbolic computations. For more advanced usage, including cloud integration and data exchange, refer to the Wolfram Client for Python documentation.

Common next steps

After successfully executing your first Wolfram Language requests, consider these common next steps to expand your capabilities:

• Explore Wolfram Language Documentation: The Wolfram Language Reference is an extensive resource covering all functions, data structures, and programming paradigms. Understanding more of the language's built-in capabilities, such as image processing, machine learning, or financial functions, will unlock more use cases.
• Utilize Wolfram Cloud: For scalable deployments, web-based interfaces, and collaborative work, explore the Wolfram Cloud. You can deploy Wolfram Language code as APIs, web forms, or interactive notebooks accessible via a web browser. This is particularly useful for exposing computational services without requiring users to install the Wolfram Engine locally.
• Advanced SDK Usage: Beyond basic evaluation, the Wolfram Client for Python supports asynchronous evaluation, data serialization, and integration with other Python libraries. Familiarize yourself with these features for more complex applications. For example, you might want to visualize Wolfram Language outputs using Python's Matplotlib library, or feed data from a Pandas DataFrame into Wolfram Language for specialized analysis.
• Integrate with External Data: Learn how to import and export data from various formats (CSV, JSON, SQL databases) into the Wolfram Language for analysis. The language has built-in functions like Import and Export that simplify these tasks Wolfram data import/export guide.
• Develop Interactive Applications: With Mathematica (the front-end for the Wolfram Language), you can create interactive applications and dashboards using functions like Manipulate and FormFunction. These are powerful for creating dynamic visualizations and user interfaces.
• Secure Cloud Deployments: When deploying APIs or web applications to the Wolfram Cloud, implement appropriate security measures. This can include setting permissions on deployed functions, using API keys for authentication, and understanding potential vulnerabilities. For general best practices in API security, consulting resources like the OAuth 2.0 specification can provide a broader context, though Wolfram's cloud authentication mechanisms might differ.

Troubleshooting the first call

Encountering issues during your initial setup or first API call is not uncommon. Here are some common problems and their solutions:

• Wolfram Engine Not Found (Command Line/Python SDK):
  • Symptom: Errors like "command not found" or "WolframKernel not found."
  • Solution: Ensure the Wolfram Engine is correctly installed and that its executable path is added to your system's PATH environment variable. On macOS, this might involve symbolic linking or adjusting your shell profile. For Python, confirm that the wolframclient package is installed and that the Python environment you are using has access to it.
• License Activation Issues:
  • Symptom: Engine prompts for activation repeatedly, or reports an invalid license.
  • Solution: Double-check your Wolfram ID and password. Ensure your internet connection is active during activation. If using a commercial license, verify that it is properly registered to your Wolfram ID and that the license key is correctly entered. Sometimes, a firewall might block the activation process; temporarily disabling it or configuring an exception might help.
• Python Session Initialization Errors:
  • Symptom: WolframLanguageSession() fails to start or reports connection errors.
  • Solution: Verify that the Wolfram Engine is runnable independently from the command line. The Python client relies on finding and launching the local kernel. Ensure no other applications are conflicting with the necessary ports if you're attempting a network-based connection (less common for local sessions). Review the Wolfram Client for Python evaluation documentation for specific known issues.
• Syntax Errors in Wolfram Language Code:
  • Symptom: The engine returns $Failed or syntax error messages.
  • Solution: Wolfram Language syntax is precise. Pay attention to square brackets [] for function arguments, curly braces {} for lists, and correct capitalization (e.g., FactorInteger, not factorinteger). Use the Wolfram Language Reference documentation to verify function names and argument structures.
• Cloud API Deployment/Access Issues:
  • Symptom: Deployed functions are inaccessible, or return permission errors.
  • Solution: Check the permissions settings on your deployed cloud object. Ensure it is set to public, or that you are providing the correct authentication if it's private. Verify the URL for the deployed function is correct. Pay attention to any rate limits or usage quotas associated with your Wolfram Cloud plan.