CPU3DAI tools for 3D, video and audio

One-2-3-45: 3D generator — what it does and what you need to run it

One-2-3-45 is an open-source 3D generator designed to quickly convert a single image into a 3D mesh. As described by the authors, it can create a 3D model from any single photo in 45 seconds without the need for per-shape optimization.

What it does

  • Generates a 3D mesh from a single input image.
  • According to the developers, the process takes 45 seconds and does not require per-shape optimization.
  • It is based on technologies related to generalizable NeRF and multi-view reconstruction.
  • The authors position the tool as a solution for 3D generation tasks and AIGC.

What you need to run it

  • Platform: the authors do not specify strict limitations; the code is written in Python.
  • Memory and GPU: specific requirements for VRAM and system RAM are not provided in the fact sheet. Given the use of NeRF and mesh generation, a discrete GPU will be required, but exact values are left to the user's discretion.
  • License: Apache-2.0. The code is open source and available in the repository.
  • Installation: local, by cloning the repository from github.com/One-2-3-45/One-2-3-45. There is no cloud version; the tool can only be run on your own hardware.

Who it suits

  • Developers and researchers in computer graphics who need fast generation of draft 3D shapes from a reference.
  • Those who are ready to set up a Python environment and have access to a GPU.
  • Users who value open source code and the ability to modify it for their own tasks.

The tool is not suitable for those looking for a ready-made cloud solution without local installation, nor for users without a suitable graphics accelerator. The authors do not provide data on minimum GPU specifications, so you should independently assess available resources before running it.

One-2-3-45 offers a research-oriented approach to the image-to-3D task, emphasizing speed and the absence of per-object fitting. The code is freely distributed, and all details about the method can be found in the repository and on the project website.

One-2-3-45 pipeline 3D generation from a single image in 45 seconds Image Single photo of an object RGB, any angle Multi-view Generates 8 views From different angles NeRF Generalizable NeRF No fine-tuning Mesh Reconstruction Multi-view stereo Final 3D model Textured mesh in 45 seconds Export formats OBJ, GLB, PLY — mesh with textures and materials Mode 1: Fast Mode 2: Accurate
How the One-2-3-45 pipeline works. The diagram is drawn from the tool’s fact sheet.

Fact sheet

Repository · Developer’s site

Code licenseApache-2.0 source
LanguagePython source
Last code change2024-04-20 source
Repository created2023-06-28 source
Changes often — as of 2026-08-03
GitHub stars1721 source
Forks103 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