AI vintage travel poster via API: 3:4 art with the headline in code

A vintage travel poster is flat color, bold shapes and a headline. On the Sume image API, 3:4 is listed by every catalog row, read 2026-10-03, so you are free to choose by price and by whether you need references. Two ways to handle the headline: let the model draw it (best on a row you have tested for lettering, like Ideogram V3 or 4.5) or leave a blank band and set it in code.

Setting it in code removes spelling risk and lets you change the place name for a series. Drawing it with the model is one step, but you must proofread each image.

Prompt skeleton

• Style: "mid-century travel poster, flat shapes, limited palette of four colors, grain".
• Scene: one landmark, one sun or moon shape, one layer of hills.
• Layout: "empty band across the bottom third, no text" for the code route.
• For the model-drawn route, quote the exact headline in double quotes and keep it to one to three words.

Which row

Generate and add the headline

Generate four options for four times the price, pick one, then add the text. A call that is still running after the 30-second wait returns 202 with a job envelope instead of 200, so check the status code before reading data; see Jobs and results.

import os
import requests

resp = requests.post(
    "https://api.sume.com/v1/images",
    headers={"Authorization": f"Bearer {os.environ['SUME_API_KEY']}"},
    json={
        "model": "qwen/qwen-image",
        "prompt": "mid-century travel poster of a lighthouse on a cliff, flat shapes, four-color palette, grain, empty band across the bottom third, no text",
        "aspect_ratio": "3:4",
        "n": 4,
    },
    timeout=60,
)
resp.raise_for_status()
if resp.status_code == 202:
    print("still running:", resp.json()["data"]["status_url"])
else:
    for image in resp.json()["data"]:
        print(image["url"])

Then set the type

Draw the headline in the empty band with Pillow, as in the book-cover recipe. Keep the poster file name the same as the place so batches stay organized.

from PIL import Image, ImageDraw, ImageFont

img = Image.open("poster.png").convert("RGB")
d = ImageDraw.Draw(img)
font = ImageFont.load_default(size=img.width // 8)
w = d.textlength("PORTO", font=font)
d.text(((img.width - w) / 2, img.height * 0.78), "PORTO", font=font, fill="#f4e8c8")
img.save("poster-porto.png")