Nano Banana 2.1 product photo 1:1: 1024, 2048, 4096 and cost per tier
Square product shots with Nano Banana 2.1: 1024, 2048 and 4096 px sizes, what each costs on Sume, and what you cannot do, such as transparent output.

A 1:1 Nano Banana 2.1 image is 1024x1024 at 1K, 2048x2048 at 2K and 4096x4096 at 4K. For a marketplace listing that asks for 2000 px or more, take 2K at $0.15 per image on Sume. Transparent cutouts are not available on this model.
Sume forwards your ratio and tier to the provider and does not document the exact pixels it returns, so this post uses Google's table. Check width and height on the first file you get back before you build a crop on them.
What do the square tiers give and cost?
Pixels come from Google's size table. The Sume column is the billed price, which is the provider list price times 1.25.
| Tier | Pixels | Google price per image | Sume billed per image |
|---|---|---|---|
| 1K | 1024x1024 | $0.0336 | $0.10 |
| 2K | 2048x2048 | $0.0504 | $0.15 |
| 4K | 4096x4096 | $0.113 | $0.20 |
Can I get a transparent background?
Not from this model on Sume. The background field with transparent is only available on ChatGPT Image 2.5 (the GPT Image 2.5 rows), and Nano Banana 2.1 rejects it with 400 unsupported_parameter. Ask for a plain white or light grey backdrop and remove it in your editor, or call a GPT Image 2.5 row when transparency is the requirement.
How do I keep the product the same across shots?
Send the product photo as a reference. On Sume, Nano Banana 2.1 accepts up to 10 references in input_references, which must be public HTTPS URLs, and aspect_ratio: "auto" on an edit matches the reference. Use a short prompt that names what must not change.
{
"model": "google/nano-banana-2.1",
"prompt": "Same bottle on a marble counter, soft side light. Keep label text unchanged.",
"input_references": [
{
"type": "image_url",
"image_url": {
"url": "https://example.com/bottle.jpg"
}
}
],
"aspect_ratio": "1:1",
"resolution": "2K"
}What should I check before listing?
Zoom to 100 percent on label text and edges, since text rendering varies per prompt. Failed generations are not billed on Sume, but a finished image with a wrong label is, so check before you batch. See the image models docs.
For a batch, fix the prompt and the reference photo, change only the setting you are testing, and run each tier once. At the Sume prices above, one image per tier costs $0.45 in total, which is a cheap way to find the lowest tier that still reads well on your marketplace. Save the winning settings as your template.
Sources
Related posts
More in Use cases
- Nano Banana 2.1 still, then Omni 1080p for 6 s: $1.225 together
A Nano Banana 2.1 still at $0.10 and a 6-second Omni Flash 1080p animation at $1.125 total $1.225 on Sume. The two-call Python script and where the cost sits.
- Narrate a Claude Motion explainer with Sume TTS and Timeline: 12 cents
A 450-character voiceover costs $0.021 on Sume TTS and a one-minute Timeline render $0.10, so narrating a Claude Motion MP4 comes to about $0.12.
- 90-Second News Explainer Video: Stills, Voiceover, Timeline, $1.06
How to build a 90-second vertical explainer on a current story with Sume: research calls, six stills, 1,200 characters of voiceover, one two-minute render.
- Nine music variants for one Reel: a flat $0.125 each, $1.125 total
Music on Sume is a flat $0.125 per generation, so nine track variants cost $1.125 whatever the length. A cost table, the request body and what to check.
Written by Sume