CPU3DAI tools for 3D, video and audio

DreamGaussian: 3D generator — what it does and what you need to run it

DreamGaussian is an open source 3D generator. The tool solves the problem of quickly creating 3D content from a text description or a single image. It is based on the Generative Gaussian Splatting method, presented by the authors at the ICLR 2024 conference.

What it does

  • Text-to-3D: generates a 3D model from a text description.
  • Image-to-3D: generates a 3D model from a single input image.
  • Representation format: uses Gaussian Splatting. The authors state that this enables efficient 3D content creation.

What you need to run it

  • Platform: local run. The code is written in Python.
  • License: MIT. Free use, modification, and distribution are permitted.
  • Repository: github.com/dreamgaussian/dreamgaussian. The latest code change is dated January 2, 2024.
  • Version: the latest release is 0.1.0. The authors do not specify exact requirements for system RAM or GPU model in the public domain.

Who it is for

The tool will be useful for researchers and developers working with generative 3D technologies who can run Python code locally. The open MIT license allows integrating the work into your own projects.

It is not suitable for those looking for a ready-made application with a GUI or a cloud service without the need to configure the runtime environment. Also note that the repository has not been updated since early 2024, and the version has early release status.

The project website is available at dreamgaussian.github.io. To assess the current state of the code and documentation, it makes sense to refer directly to the repository.

DreamGaussian pipeline Generative Gaussian Splatting for 3D content creation INPUT Image or text image-to-3d text-to-3d INITIALIZATION Point cloud generation Gaussian Splatting initial geometry OPTIMIZATION Refining shape and color iterative rendering differentiable RESULT Ready 3D model Gaussian Splatting efficient rendering Text Img. Stage 1 Stage 2 Stage 3 Stage 4 Open source (MIT) • Python • github.com/dreamgaussian/dreamgaussian ICLR 2024 Oral • Generative Gaussian Splatting for Efficient 3D Content Creation Export formats: PLY • OBJ • GLB
How the DreamGaussian pipeline works. The diagram is drawn from the tool’s fact sheet.

Fact sheet

Repository · Developer’s site

Code licenseMIT source
LanguagePython source
Last code change2024-01-02 source
Repository created2023-09-27 source
Changes often — as of 2026-08-03
Latest release0.1.0 source
Release date2023-10-21 source
GitHub stars4350 source
Forks399 source

Values are collected automatically from official sources and were checked on 2026-08-03. Each one links to its source, and values taken from the developer’s pages also carry a verbatim quote — hover over the note. Pricing and versions are shown as of the check date and change most often; verify on the vendor’s site before buying.

See also