FLUX infographic prompts: layout, icons, palette, one Sume request

Black Forest Labs says to specify the layout, content hierarchy, icons, and color palette in a FLUX infographic prompt. Write those four things as plain sentences in prompt and send them to a FLUX model id such as black-forest-labs/flux.2-pro through Sume's image route, with aspect_ratio set for the shape you want.

The guidance is from BFL's infographics page, read 2026-10-01; the page's own advice is to be explicit about structure (steps, sections, columns) and visual style (minimalist, colorful, corporate).

What goes in the prompt?

One sentence per element keeps the prompt checkable. Layout: how many sections and in which direction. Hierarchy: the title, then each heading and its one line. Icons: what each section shows. Palette: two or three named colors on a plain background. For a longer walkthrough, see AI infographic generator.

What does the Sume request look like?

curl -X POST "https://api.sume.com/v1/images" \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "black-forest-labs/flux.2-pro",
    "prompt": "Infographic, three columns with steps numbered left to right. Title on top, then a heading and one short line per step. A simple line icon above each heading. Navy and orange on a white background.",
    "aspect_ratio": "4:3",
    "n": 2
  }'

Which request fields matter?

Per the image docs, aspect_ratio takes values such as 1:1, 16:9, 9:16, 4:3, and 3:4, and n is 1 to 10. A parameter the chosen model does not list is rejected with 400 unsupported_parameter, so check GET /v1/images/models for the model you pick. Custom pixel image_size is accepted on models that take custom pixels, which include Flux (Image 1.0).

How do I check the text in the result?

Read every label in the output before use; generated lettering can contain mistakes. Rerun with n greater than 1 and keep the cleanest. See FLUX.2 prompt guide.