n8n Edit Image text alignment over a Sume-generated image

Generate the image with Sume first, then let n8n's Edit Image node write the words on it. Its Add Image Text operation now has horizontal and vertical alignment options, so you can center a caption without hand-tuning X and Y. Sume supplies the picture; n8n places the text.

The alignment change is from the n8n release notes, read 2026-09-30. The job and result behavior is from Sume's Jobs and results page.

What did n8n 2.39 change?

n8n 2.39 (2026-09-08) says the Edit Image node's Add Image Text operation now includes Horizontal Alignment (Left, Center, Right) and Vertical Alignment (Top, Middle, Bottom), letting you position text without manual trial and error on Position X/Y.

How do the two steps connect?

Submit the Sume request in async mode. The response carries request_id (the job id), status_url and result_url. Results are available only after completion, so poll status_url until the job is terminal before asking for the result. A completed job lists artifacts, each with a url on media.sume.com; use those Sume media URLs, not raw provider URLs. Download that URL into the workflow (my suggestion: an HTTP Request node returning a file), then pass the binary to Edit Image.

Why overlay text instead of prompting for it?

When the words must be exact, a title, a price or a legal line, writing them in your workflow keeps them under your control and lets you change them without regenerating the picture. That is a design choice, not a Sume or n8n claim. Prompt for an image with clear empty space where the text will sit.

What about the wait?

Do not hold one HTTP node open for a slow job; poll instead, as covered in n8n Wait node and next_poll_after_seconds. For the basic generation setup see n8n AI image generation.