OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Developer Tools & SDKs

blendtorch

blendtorch is a Python framework that connects Blender with PyTorch data pipelines, using Eevee to generate synthetic images and annotations for deep-learning workflows. It supports real-time streaming of distributed renderings, recording for replay without Blender, and remotely controlled Blender environments for reinforcement-agent training.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

blendtorch — OpenViST project profile in AI Developer Tools & SDKs.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use blendtorch when a PyTorch workflow needs Blender-generated images and annotations, replayable synthetic data, or a Blender simulation connected to reinforcement-agent training.

Who it is for

  • Deep-learning practitioners generating synthetic visual training data.
  • Researchers using domain randomization or supervised learning with Blender scenes.
  • Reinforcement-learning developers who want Blender to serve as a simulation and visualization environment.

Documented uses

  • Generate synthetic images and annotations for model training.
  • Stream Blender renderings into PyTorch data pipelines.
  • Record generated data for replay without Blender.
  • Train reinforcement agents in remotely controlled Blender environments.
Before you choose

Requirements & limits

The project repository is public and declares an MIT license.

  • The Blender and PyTorch components install into separate Python environments.
  • Blender’s executable must be on PATH for the documented installation process.
  • The listed tested operating systems are Windows 10 and Linux; the project notes that other versions may work but are not confirmed as tested.
  • OpenAI Gym is optional, but advised for reinforcement-learning use.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install blendtorch’s Blender and PyTorch components

Prepare the two blendtorch sub-packages in their respective Python environments.

Before you start

  • A cloned blendtorch repository
  • Blender and PyTorch installed
  • Blender executable available on PATH
  • Blender opened at least once to complete initial settings

Steps

  1. Clone the repository.

  2. Add Blender’s executable to PATH.

  3. Run blender --background --python /scripts/install_btb.py to install the Blender-side component in Blender’s bundled Python environment.

  4. Run the command below in the Python environment intended for PyTorch.

    Documented command
    pip install -e /pkg_pytorch

Expected outcome

The Blender-side and PyTorch-side blendtorch components are installed in their respective environments.

Guide limitations

  • The instructions list tested Blender versions 2.83, 2.91, 3.0 and 3.1, and PyTorch versions 1.5 and 1.10; other versions may work but are not documented as tested.
  • OpenAI Gym is optional and advised for reinforcement-learning use.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
cheind/pytorch-blender · Official homepage title · retained in its source language
How we review sourcesSuggest a correction