NS WaveSpeed Nano Banana Pro Text to Image Multi

The "Multi" variants in the Symbiotica WaveSpeed family all have the same job: give you more than one take per queue run. This one is the text-to-image version - one prompt, two Nano Banana Pro images, in a single call. If you've ever queued the same prompt twice and kept the better one, this is that habit turned into a node.

It's part of the Symbiotica pack's neuralsins/WaveSpeed family. It's a thin node, and that's fine - it exists to save you a round trip, and the batch output drops straight into a compare-grid or a Pick-style workflow where you eyeball candidates and keep one.

How it works

prompt in, two images out of output_images (IMAGE) as a batch - index 0 and 1, same aspect ratio, different draws. It POSTs to WaveSpeed's Nano Banana Pro text-to-image path and asks for two results.

The inputs you set:

• prompt - the image description. Both images follow it.
• aspect_ratio - a slim 4-choice set: 3:2 (default), 2:3, 3:4, 4:3. No 16:9, no 1:1 - if your deliverable needs those, the single-output Pro T2I node has the full ten-ratio menu.
• num_images - fixed at 2. The tooltip says so. Don't hunt for a way to raise it.
• output_format - png or jpeg.

enable_sync_mode defaults on, so you wait inline and get both tensors when it's done.

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), then wire an NS WaveSpeed Client with your WAVESPEED_API_KEY into client.

Common issues

• Two images, one bill. This generates two at once, so the call costs about double the single-output version. You're paying for convenience, not getting a discount.
• No resolution dropdown. Unlike the single-output Pro T2I with its 1k/2k/4k ladder, this variant runs at the model's default and offers no resolution control.
• Aspect ratio is limited to four. Plan the shape before you queue, or you'll be re-rolling.
• No seed. Reproducibility isn't on the menu here either - same as the rest of the Pro Multi family.

The honest read: reach for this when you know you'll want candidates to choose between. For a single, carefully-prompted image where you already know what you want, the single-output node is the leaner call.