ComfyUI Node
Gemini Spatial Understanding
A ComfyUI node in Gemini/Spatial with 8 inputs and 2 outputs.
Gemini Spatial Understanding
- image
- annotated_image
- json_output
◄modelgoogle/gemini-2.5-flash►
◄task_type2D bounding boxes►
◄targetitems►
◄temperature0.0►
◄reasoning_effortnone►
◄service_tierdefault►
◄api_key►
CategoryGemini/Spatial
Inputs (8)
| Name | Type | Default | Description |
|---|---|---|---|
| image | IMAGE | — | |
| model | STRING | google/gemini-2.5-flash | — |
| task_type | COMBO | 2D bounding boxes | 2 options: 2D bounding boxes, Points |
| target | STRING | items | — |
| temperature | FLOAT | 0.00–2 | — |
| reasoning_effort | COMBO | none | 6 options: none, minimal, low, medium, high, xhigh |
| service_tier | COMBO | default | 3 options: default, flex, priority |
| api_key | STRING | — |
Outputs (2)
| Name | Type | Description |
|---|---|---|
| annotated_image | IMAGE | — |
| json_output | STRING | — |