MoGe is a powerful model for recovering 3D geometry from monocular open-domain images, including metric point maps, metric depth maps, normal maps and camera FOV. Check our websites (MoGe-1, MoGe-2, MoGe-3) for videos and interactive results!

Overview

• Accurate 3D geometry estimation: Estimate point maps & depth maps & normal maps from open-domain single images with high precision -- all capabilities in one model, one forward pass.
• Optional ground-truth FOV input: Enhance model accuracy further by providing the true field of view.
• Flexible resolution support: Works seamlessly with various resolutions and aspect ratios, from 2:1 to 1:2.
• Optimized for speed: Achieves 60ms latency per image (A100 or RTX3090, FP16, ViT-L). Adjustable inference resolution for even faster speed.

(2026-08-18)

• ❗Released MoGe-3, featuring significantly improved fine-grained point map geometry.

Requires Python 3.10 or newer. Dependencies are declared in pyproject.toml, which works with both uv and pip.

macOS is not supported: MoGe-3 depends on FlexGEMM, which builds on Triton, and Triton publishes no macOS wheels.

The following optional extras are available:

git clone https://github.com/microsoft/MoGe.git
cd MoGe
uv sync                             # inference, all model versions
# uv sync --extra train             # ... plus the training dependencies