Ideogram layerize API: text blocks vs Sume's flat image

Ideogram's POST /v2/design/layerize/ideogram-3 turns a flat image into an editable design: detected text comes back as positioned text_blocks with matched fonts, sizes and colors, next to a text-free base image. Sume's images API does not return layers. A generation returns data[].url, a Sume-hosted signed URL to one flat image.

Ideogram facts are from its API reference; Sume facts from Image generation, read 2026-10-01.

What does layerize return?

The request is multipart with a required image (max 50MB), an optional prompt to guide text detection, a seed, and up to 5 font_candidate_files (.ttf, .otf, .woff, .woff2) for style matching. Results come back directly by default, or with async or a webhook_url as a generation_id to poll. Each LayerizedImageObject carries resolution, text_blocks, is_image_safe and a url; the reference notes links are available for a limited time, so download what you keep.

How do the outputs differ?

What can I do on Sume instead?

Regenerate or edit the image with a prompt, and handle text overlay in your own design tool. The docs publish parameters per model as capability descriptors, and a request that sets a parameter the model does not list is rejected, so a layers or text_blocks style field will fail rather than be ignored. For a sibling comparison see Seedream 5.0 Pro layers versus Sume's flat result.

Should I expect editable fonts from Sume?

No. Nothing in the images docs returns font names, sizes or colors. If you need editable type, keep your own text layer and use a Sume image as the background.