ComfyUI Node
Bobs LoRA Loader (Universal)
Block-weighted LoRA loading for any architecture ComfyUI supports — SD1.5, SD2, SDXL, SD3/3.5, FLUX, Chroma, AuraFlow, PixArt, HiDream, Qwen-Image, Wan, LTX-Video, Mochi, HunyuanVideo/DiT, Lumina, Cosmos and others. The block stacks are discovered from the loaded model itself rather than a hard-coded table, so new and pruned architectures work without an update. Weights are assigned by depth: Early through Late across the whole block stack, plus embeddings, output head and text encoder.
Bobs LoRA Loader (Universal)
- model
- clip
- MODEL
- CLIP
- info
◄lora_name▾►
◄strength1.00►
◄preset▾►
◄Text Encoder1.00►
◄Input & Embeddings1.00►
◄Early Blocks (Composition)1.00►
◄Early-Mid Blocks (Subject)1.00►
◄Mid Blocks (Concept & Style)1.00►
◄Late-Mid Blocks (Detail)1.00►
◄Late Blocks (Texture)1.00►
◄Output Head1.00►
◄Other Tensors1.00►
CategoryBobs_Nodes
Inputs (14)
| Name | Type | Default | Description |
|---|---|---|---|
| model | MODEL | Diffusion model to patch. | |
| lora_name | COMBO | LoRA file to load. 'None' passes the model through untouched. | |
| strength | FLOAT | 1.00-5–5 | Global multiplier applied on top of every block weight. |
| preset | COMBO | Choose 'Custom' to use the sliders below; any other preset overrides them. | |
| Text Encoder | FLOAT | 1.00-2–2 | All text-encoder weights (CLIP / T5 / LLM), whatever the model uses. |
| Input & Embeddings | FLOAT | 1.00-2–2 | Patch, timestep, guidance and context embedding layers feeding the block stack. |
| Early Blocks (Composition) | FLOAT | 1.00-2–2 | First 20% of the block stack. Global composition and layout. |
| Early-Mid Blocks (Subject) | FLOAT | 1.00-2–2 | 20–40% of the block stack. Subject identity. |
| Mid Blocks (Concept & Style) | FLOAT | 1.00-2–2 | 40–60% of the block stack. Concept and dominant style. |
| Late-Mid Blocks (Detail) | FLOAT | 1.00-2–2 | 60–80% of the block stack. Detail generation. |
| Late Blocks (Texture) | FLOAT | 1.00-2–2 | Final 20% of the block stack. Fine texture and surface. |
| Output Head | FLOAT | 1.00-2–2 | Final projection back to latent space. |
| Other Tensors | FLOAT | 1.00-2–2 | Tensors that matched no stack, embedding or head pattern. |
| clipopt | CLIP | Optional. Leave unconnected to patch the model only. |
Outputs (3)
| Name | Type | Description |
|---|---|---|
| MODEL | MODEL | Model with the block-weighted LoRA applied. |
| CLIP | CLIP | CLIP with the text-encoder portion of the LoRA applied. |
| info | STRING | Per-block report: weight, tensors found and tensors actually patched. |