NS WaveSpeed Nano Banana Pro Edit Multi
Two edits at once, one prompt — Pro Edit Multi
- client
- output_images
Same engine as Nano Banana Pro Edit, different output philosophy: instead of one edited image, this node gives you two takes on the same instruction in a single call. Same prompt, same references, two interpretations - which is exactly what you want when you're iterating on a look and don't want to re-queue for every variant.
It's part of the Symbiotica pack's neuralsins/WaveSpeed family, and it's a thin node even by that family's standards. The num_images input is fixed at 2 - the tooltip says so outright - so what you're really choosing is "batch of two" vs. "batch of one, per call." If you find yourself running the Pro Edit node twice and keeping one, this node is the cheaper way to do it in one round trip.
How it works
You give it a prompt and up to six image URLs (image_1 required, image_2–image_6 optional references), and it POSTs to WaveSpeed's Nano Banana Pro edit endpoint asking for two results. Both come back stacked in output_images (IMAGE) as a batch - index 0 and 1, same size, same aspect, different results. A batch-index or Split Image node downstream will pull them apart if you want to compare side by side.
The inputs worth knowing:
prompt- the edit instruction. Both variants follow it.image_1…image_6- image URLs; only the first is required, the rest are reference material.aspect_ratio- a slimmer set than the single version:3:2(default),2:3,3:4,4:3. Note there's noresolutiondropdown here - this variant runs at the model's default rather than offering the 1k/2k/4k ladder.output_format-pngorjpeg.
enable_sync_mode defaults on.
Installing
cd ComfyUI/custom_nodes
git clone https://github.com/symbiotica-ai/comfyui-nodes.git symbiotica
pip install -r symbiotica/requirements.txt
Restart or install via ComfyUI Manager (search Symbiotica). Wire an NS WaveSpeed Client with your WAVESPEED_API_KEY into client.
Common issues
- It's two per call, always. Don't hunt for a count control -
num_imagesexists but is fixed at 2. Want four? Run it twice. - Two variants, one bill. This costs more per call than the single-output edit because it generates two images. That's the trade - you're buying a batch, not a discount.
- References are URLs, not tensors. The standing rule for this family.
- Aspect ratio surprises. The limited 4-choice set means no 16:9 here - if your deliverable is widescreen, this variant can't make it, and you want the single-output Pro Edit with its ten-ratio menu instead.
The honest read: this is a convenience node. It saves you a queue round-trip when you know you'll want a second take, and the batch output slots neatly into the compare-grid nodes. Beyond that, it does exactly what it says and nothing more.
Inputs (12)
| Name | Type | Default | Description |
|---|---|---|---|
| client | WAVESPEED_AI_API_CLIENT | — | |
| prompt | STRING | Text description of the edit to perform | |
| 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 | 3:2 | Aspect ratio of the output images |
| num_images | INT | 22–2 | Number of edited images to generate (fixed at 2) |
| 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_images | IMAGE | — |