Evaluate & Install

HOOPS AI Package & License Key

HOOPS AI comes with a 60-day evaluation license. During your evaluation, a dedicated Consulting Engineer will be assigned to assist you. For technical support or to extend your evaluation license, please contact your assigned Consulting Engineer.

Tip

Start your Trial Experience to get immediate access to HOOPS AI.

Installing HOOPS AI

Already have Conda installed? Get started in just two commands:

pip install hoops-ai-installer --extra-index-url https://packages.techsoft3d.com/pip
hoops-ai-installer install --mode cpu --extra-index-url https://packages.techsoft3d.com/pip

Prerequisites and Environment Setup

Install Miniconda

Miniconda is a lightweight installer for Conda, a package and environment manager. Follow these steps to install:

  1. Download the latest Miniconda for Windows by clicking here.

    Miniconda installer on Windows
  2. Run the installer and follow the prompts.
    Note: Make sure to add Miniconda to your PATH environment variable.
    Miniconda environment setup on Windows
    Double-check that miniconda and condabin are in the path: i.e:
    C:\Users\**{UserName}**\miniconda3
C:\Users\**{UserName}**\miniconda3\condabin
    Windows Path environment variable

  3. Verify the installation by opening PowerShell and running:

    conda --version
python --version

    These commands display the installed versions of conda and Python, confirming that Miniconda is properly installed.

    Miniconda installation success on Windows

Note

Make sure conda is in your PATH. On Windows, verify that C:\Users\{USERNAME}\miniconda3\Scripts is in your path. On Linux, ensure ~/miniconda3/bin is in your PATH (the installer adds this automatically if you accepted conda init).

Note

Conda 25+ (first-time setup): Starting with Conda version 25, you will be prompted to accept the Anaconda Terms of Service the first time you run a conda command. This is expected and only needs to be done once. Simply follow the prompt in your terminal to accept the terms and continue.

Install CUDA (Required for GPU Support)

CUDA 11.7 is only required if your workflow needs nvcc or other toolkit binaries. For runtime execution with PyTorch GPU wheels, the NVIDIA driver (515.48+) is the hard requirement.

Follow these steps to install the CUDA toolkit:

  1. Download CUDA 11.7 from the NVIDIA CUDA Toolkit Archive.

  2. Select your OS-specific installer and download.
    Run the installer and follow the prompts.
  3. Ensure the CUDA installation path is included in your system’s PATH environment variable.
    The default installation path is typically:
    C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.7\bin
C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.7\libnvvp

  4. Verify the installation by opening PowerShell and running:

    nvcc --version

    This command displays the installed version of CUDA, confirming that CUDA is properly installed.

Once you fill out the form to evaluate HOOPS AI, one of our Sales Representatives will get in touch with you to discuss your needs and objectives. If they align with our policy, you will receive an email with a download link and installation instructions for HOOPS AI.

Step 1: Download the installer

Download the installer from our private artifactory using the complete pip install command provided in this download page.

# Example command - use the complete command from your email
pip install hoops-ai-installer --extra-index-url https://packages.techsoft3d.com/pip

Step 2: Run the installer

Run from any directory:

CPU version:

hoops-ai-installer install --mode cpu --extra-index-url https://packages.techsoft3d.com/pip

GPU Version (CUDA 11.7 required)

hoops-ai-installer install --mode gpu --extra-index-url https://packages.techsoft3d.com/pip

The installer will:

  • ✅ Check prerequisites (Python 3.9, Conda, CUDA for GPU)

  • ✅ Create conda environment from environment_cpu.yml or environment_gpu.yml

  • ✅ Install all dependencies (PyTorch, DGL, ML libraries, Jupyter, etc.)

  • ✅ Install HOOPS AI, Exchange, and Viewer packages

  • ✅ Register Jupyter kernel

  • ✅ Setup license (if license.env exists)

  • ✅ Setup converter path (if packages/ exists)

  • ✅ Verify installation

Activate The Environment

To start using HOOPS AI, activate the Conda environment created by the installer:

# Activate the environment
conda activate hoops_ai_cpu      # For CPU version
conda activate hoops_ai_gpu      # For GPU version

# Launch JupyterLab
jupyter lab

⚠️ First-time conda users: If you get CondaError: Run 'conda init' before 'conda activate'

  1. Run the initialization command:

    Before you can activate environments in PowerShell, you must initialize conda:

    conda init powershell

    This command configures PowerShell to recognize conda activation commands. You must close and reopen PowerShell for the changes to take effect.

  2. After initialization and reopening PowerShell, activate the environment:

conda activate hoops_ai_cpu      # For CPU version
conda activate hoops_ai_gpu      # For GPU version

This switches to the newly created environment, making its packages and Python interpreter active. When successful, you will see (hoops_ai_cpu) prepended to your command prompt, indicating that all commands are now executed within this conda environment.

To verify the environment was created and list all available environments:

conda info --envs

This displays all conda environments on your system. At this stage, the new environment exists but is not yet active. (the active environment marked by an asterisk (*)).

Note

In JupyterLab, ensure you select the kernel named HOOPS AI (CPU) or HOOPS AI (GPU) to run notebooks within the correct environment.

Troubleshooting

“CondaError: Run ‘conda init’ before ‘conda activate’”

Most common issue for new conda users!

This means conda hasn’t been initialized for your shell yet.

Solution:

  1. Run conda init bash (Linux), conda init powershell (PowerShell), or conda init (Command Prompt)

  2. Close and completely reopen your terminal (this step is required!)

  3. Now conda activate hoops_ai_cpu will work

Alternative (Windows): Use the full activation path:

C:\Users\YourUsername\miniconda3\Scripts\activate hoops_ai_cpu

Alternative (Linux): Use the full activation path:

source ~/miniconda3/bin/activate hoops_ai_cpu

“Python 3.9–3.12 supported”

  • You can have any local Python version from 3.9 to 3.12 installed python.org.

  • The project runs inside a Conda environment that will use Python 3.9.

“Conda not found”

Install Miniconda and ensure it’s in PATH. Restart terminal after installation.

On Windows, run from Anaconda/Miniconda Prompt where conda is pre-configured, or add conda to PATH.

On Linux, ensure ~/miniconda3/bin is in your PATH, or initialize conda for your shell:

conda init bash
# Then restart your terminal

To verify conda is available:

conda --version
which conda          # Linux
where conda          # Windows

“CUDA 11.7 not found” (GPU)

Install CUDA 11.7 from NVIDIA CUDA Toolkit Archive.

On Linux, nvcc may not be installed by default. This is optional for runtime-only use : nvidia-smi must work (driver installed). If your workflow needs toolkit binaries, install CUDA 11.7 and verify with nvcc --version.

GPU on Linux: Environment File Compatibility

If installation fails while solving conda packages on Linux, the bundled GPU YAML may contain Windows-only packages. The installer automatically prefers environment_gpu_linux.yml on non-Windows systems when present.

If needed, you can also create the environment manually:

conda env create -f hoops_installer/environment_gpu_linux.yml -n hoops_ai_gpu

“CUDA available: False” after GPU installation

If torch.cuda.is_available() returns False:

  • Outdated NVIDIA driver : Update to 515.48 or newer

  • GPU not CUDA-compatible : Check your GPU’s compute capability (3.5+ required)

  • CUDA installation incomplete : Reinstall CUDA 11.7

Verify with:

nvidia-smi          # Should show your GPU
nvcc --version      # Should show CUDA 11.7 (optional)

Conda Terms of Service Error

If you see CondaToSNonInteractiveError: Terms of Service have not been accepted:

Option 1: Upgrade hoops-ai-installer (recommended):

pip install --upgrade hoops-ai-installer

Option 2: Accept Anaconda TOS (quick fix):

conda tos accept --override-channels --channel https://repo.anaconda.com/pkgs/main
conda tos accept --override-channels --channel https://repo.anaconda.com/pkgs/r
conda tos accept --override-channels --channel https://repo.anaconda.com/pkgs/msys2

Unicode/Emoji Display Errors

If you see UnicodeEncodeError: 'charmap' codec can't encode character '\u2705' during verification, this is just a cosmetic issue with emoji display in Command Prompt. The installation succeeded. Use PowerShell or Windows Terminal for better Unicode support.

Import Errors

conda deactivate
conda activate hoops_ai_cpu

Environment Already Exists

The installer will ask if you want to recreate it. Choose “yes” to update.

Getting Started with Tutorials

HOOPS AI includes hands-on tutorials to help you learn by doing. These tutorials use the same conda environment you just installed.

Tutorial Notebooks

Access our interactive Jupyter notebooks from the public GitHub repository:

Clone the repository or download the notebooks directly to get started with step-by-step examples.

You can also find the tutorials in this documentation: Tutorials

Tutorial Data Assets

The tutorials require sample CAD files and datasets. Download the tutorial data package using the link provided in your Sales team email, or access it directly:

After downloading, extract the data to a convenient location and update the file paths in the tutorial notebooks accordingly.

Note

The tutorial notebooks are designed to run in the hoops_ai_cpu or hoops_ai_gpu environment you created during installation. Make sure to activate the environment and select the correct Jupyter kernel before running the tutorials.

Next Steps

Now that you have Downloaded the Package and know where to get an extended evaluation or perpetual license key you are ready to read the following pages :

Ready? Simply press the button Next below to get started!