ComfyUI Node
Depth Anything Tensorrt Engine Builder ⚡
Downloads a Depth Anything ONNX model and converts it to a TensorRT engine optimized for your GPU. The engine only needs to be built once and is reused for all subsequent inference.
Depth Anything Tensorrt Engine Builder ⚡
- message
◄model_versionv1_large►
◄custom_engine_name►
◄use_fp16true►
◄custom_onnx_path►
Categorytensorrt
Inputs (4)
| Name | Type | Default | Description |
|---|---|---|---|
| model_version | COMBO | v1_large | Select the Depth Anything model version and size. Larger models provide better quality but require more VRAM. |
| custom_engine_name | STRING | Optional custom name for the TensorRT engine file. If empty, will use the default name based on the model. | |
| use_fp16 | BOOLEAN | true | Enable FP16 precision for faster inference and lower VRAM usage. Disable if you experience stability issues. |
| custom_onnx_path | STRING | Optional path to a custom ONNX model file. If provided, will use this instead of downloading the predefined model. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| message | STRING | Status message indicating whether the engine was built successfully, already exists, or encountered an error. |