ComfyUI Node
DepthAnything3 (CRT)
A ComfyUI node in CRT/DepthAnything3 with 12 inputs and 1 output.
DepthAnything3 (CRT)
- image
- image
◄model_namedepth_anything_3_small.safetensors►
◄keep_model_loadedtrue►
◄weight_dtypedefault►
◄megapixels0.50►
◄modemono►
◄outputdepth►
◄normalizationv2_style►
◄apply_sky_clipfalse►
◄coloredfalse►
◄ref_view_strategysaddle_balanced►
◄pose_methodcam_dec►
CategoryCRT/DepthAnything3
Inputs (12)
| Name | Type | Default | Description |
|---|---|---|---|
| model_name | COMBO | depth_anything_3_small.safetensors | 4 options: depth_anything_3_small.safetensors, depth_anything_3_base.safetensors, depth_anything_3_mono_large.safetensors, depth_anything_3_metric_large.safetensors |
| keep_model_loaded | BOOLEAN | true | — |
| weight_dtype | COMBO | default | 4 options: default, fp16, bf16, fp32 |
| image | IMAGE | — | |
| megapixels | FLOAT | 0.500.05–8 | — |
| mode | COMBO | mono | 2 options: mono, multiview |
| output | COMBO | depth | 4 options: depth, depth_colored, sky_mask, confidence |
| normalization | COMBO | v2_style | 3 options: v2_style, min_max, raw |
| apply_sky_clipopt | BOOLEAN | false | — |
| coloredopt | BOOLEAN | false | — |
| ref_view_strategyopt | COMBO | saddle_balanced | 4 options: saddle_balanced, saddle_sim_range, first, middle |
| pose_methodopt | COMBO | cam_dec | 2 options: cam_dec, ray_pose |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| image | IMAGE | — |