MoGe: Monocular Geometry Estimation
ComfyUI Wiki
MoGe by Microsoft provides accurate monocular geometry estimation for ComfyUI, with v1 and v2 ViT-L variants including normal prediction.
MoGe is a monocular geometry estimation model from Microsoft Research. It predicts high-quality depth maps and surface normals from a single input image. The v2 variant adds improved normal prediction capabilities. Useful for 3D-aware image editing, ControlNet conditioning, and scene reconstruction in ComfyUI.
Usage in ComfyUI
Place the model files in ComfyUI/models/geometry_estimation/. Requires a ComfyUI custom node that supports MoGe for loading and inference.
| Developer | Microsoft Research |
| Release Date | 2025 |
| Architecture | ViT-L based Geometry Estimation |
| License | MIT |
| Type | Geometry Estimation / Depth + Normal |
| Variants | v1 ViT-L fp16, v2 ViT-L fp16 (normal) |
Links
- Hugging Face (Comfy-Org): Comfy-Org/MoGe
- Upstream: Ruicheng/moge-vitl
- GitHub: microsoft/moge
Guides and workflows related to this model series.
No articles found.
Comments
Sign in with GitHub to join the discussion.