1:8 vertical strip split into four 1:2 panels: cost vs four calls

A 1:8 image cut into four equal pieces gives four 1:2 panels, the shape a long scroll or vertical story sequence needs. Nano Banana 2.1 lists 1:8 on Sume at $0.15 for 2K; four separate 1:2 images on models that list that ratio cost $0.10 to $0.15, but they will not share one continuous scene.

The math

For a width w the 1:8 image is 8w tall. Four panels are 2w tall each, so each panel is w by 2w, which is 1:2. Eight panels would be 1:1.

Which route to use

The cheapest totals belong to four separate calls on models that list 1:2, but each panel is its own scene. If the story needs continuity across panels (one road, one skyline, one gradient), the single 1:8 call gives it to you. Grok Image and Flux 2 Pro list 1:2 in the repo catalog; confirm in GET /v1/images/models before you send.

import os, requests

body = {
    "model": "google/nano-banana-2.1",
    "prompt": "A long vertical scroll from sky to ocean floor, one continuous scene, no text",
    "aspect_ratio": "1:8",
    "resolution": "2K",
}
r = requests.post(
    "https://api.sume.com/v1/images",
    headers={"Authorization": f"Bearer {os.environ['SUME_API_KEY']}"},
    json=body, timeout=60,
)
print(r.status_code)  # 200 = image body, 202 = job envelope
if r.status_code == 200:
    out = r.json()
    print(out["data"][0]["url"], out["usage"]["cost"])

Cut into four

Cut at equal heights with Pillow.

from PIL import Image
img = Image.open("strip.png")
w, h = img.size
panel = h // 4
for i in range(4):
    img.crop((0, i * panel, w, (i + 1) * panel)).save(f"panel-{i + 1}.png")

Gotchas

• Nano Banana Pro does not list 1:8; among the Nano Banana rows only 2.1 does.
• A tall 4K call may return 202; poll the job and read the result.
• Ask for calm detail near the three cut lines.