ComfyUI Node
🧠 DeepSeek OCR (Images → Markdown)
A ComfyUI node in SwanOCR with 3 inputs and 1 output.
🧠 DeepSeek OCR (Images → Markdown)
- images
- markdown_text
◄modeGundam►
◄task_typedocument►
CategorySwanOCR
Inputs (3)
| Name | Type | Default | Description |
|---|---|---|---|
| images | IMAGE | — | |
| mode | COMBO | Gundam | 5 options: Tiny, Small, Base, Large, Gundam |
| task_type | COMBO | document | 5 options: document, without layouts, other image, figures in document, general |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| markdown_text | STRING | — |