NS WaveSpeed Nano Banana Pro Edit
Nano Banana Pro edits with six reference images
- client
- output_image
Nano Banana Pro is Google's flagship image model - 4K native output, and it reasons about your prompt before it renders. This node is WaveSpeed's hosted version of its edit mode, with room for six reference images. That reference budget is the whole story: it's what turns "edit this image" into "edit this image to look like these people, this style, and this setting."
It's part of the Symbiotica pack's neuralsins/WaveSpeed family, and it sits a step above the plain Nano Banana edit node. The KB's closed-source doc puts Pro's pricing around $0.039–0.24 per image at Google's own rates; WaveSpeed's tooltip quotes its own meter right on the node - 1k and 2k at $0.14, 4k at $0.24. No open weights, ever - this only exists as an API call.
How it works
The prompt is your edit instruction. image_1 is the primary image; image_2 through image_6 are optional references the model can consult - a face to match, a style to follow, a color palette to respect. It POSTs to WaveSpeed's /api/v3/google/nano-banana-pro/edit and returns the edited result through output_image (IMAGE).
The inputs that matter:
prompt- describe the edit precisely. Pro's "thinking mode" rewards a well-specified sentence.image_1…image_6- image URLs. Onlyimage_1is required; the rest are optional references. Wire in a subject reference and the model holds identity far better than a bare edit.aspect_ratio- ten choices, 1:1 through 21:9, default 1:1. Pick your deliverable shape up front.resolution-1k,2k, or4k, default 1k. 4k is the reason to reach for Pro over the base model.output_format-png(default, transparency) orjpeg.
enable_sync_mode defaults on - the node waits and returns the finished tensor.
Installing
cd ComfyUI/custom_nodes
git clone https://github.com/symbiotica-ai/comfyui-nodes.git symbiotica
pip install -r symbiotica/requirements.txt
Restart (or ComfyUI Manager → Symbiotica). An NS WaveSpeed Client node with your WAVESPEED_API_KEY feeds the client socket - settings key, env var, or directly on the node.
Common issues
- Six image slots, one is required. You can leave 2–6 empty; they're optional references, not mandatory inputs. Empty slots are simply skipped.
- References are URLs. Same family rule - you can't wire local tensors in. Host them or use a URL-emitting loader.
- Identity still drifts. Pro holds a face better than the base model, but with multiple references it can blend features. Fewer, well-chosen references beat more conflicting ones.
- 4k doubles the price and the wait. The tooltip is honest about it - decide whether the deliverable actually needs it.
The trap to avoid is assuming Pro is strictly better because it's bigger. For a fast edit with one reference, the base Nano Banana node is cheaper. Pro earns its premium when you're juggling references or need 4K.
Inputs (12)
| Name | Type | Default | Description |
|---|---|---|---|
| client | WAVESPEED_AI_API_CLIENT | — | |
| prompt | STRING | Text description of the edit to perform (e.g., 'Replace the cloudy sky with a clear sunset') | |
| image_1 | STRING | Primary input image URL (connect from Upload Image node) | |
| image_2 | STRING | Second input image URL (optional) | |
| image_3 | STRING | Third input image URL (optional) | |
| image_4 | STRING | Fourth input image URL (optional) | |
| image_5 | STRING | Fifth input image URL (optional) | |
| image_6 | STRING | Sixth input image URL (optional) | |
| aspect_ratio | COMBO | 1:1 | Aspect ratio of the output image |
| resolution | COMBO | 1k | Output resolution: 1k ($0.14), 2k ($0.14), 4k ($0.24) |
| output_format | COMBO | png | Output format - use PNG for transparency support |
| enable_sync_mode | BOOLEAN | true | Wait for generation to complete before returning |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| output_image | IMAGE | — |