Ideogram 4.5 size source: keep dimensions, and Sume aspect auto
Ideogram 4.5 size takes auto, source or WIDTHxHEIGHT; source needs source images. Sume's size is a tier shorthand; use aspect_ratio auto to match a reference.

Ideogram 4.5's size accepts "auto", "source" or an exact WIDTHxHEIGHT; source is rejected when you send no source images. Sume has no source value: its size is only a shorthand for a resolution tier, explicit pixels are not served in v1, and the way to match a reference is aspect_ratio: "auto".
Ideogram facts are from its Ideogram 4.5 reference; Sume facts from the Image API, read 2026-10-01.
What does each Ideogram 4.5 size value do?
Without source images, an exact size must be one of the supported 1K or 2K presets (the page gives 1024x1024, 2048x2048 and 1440x2880 as examples), auto or an omitted field picks a supported size from the prompt, and source is rejected. With source images, auto is the default and picks a supported 2K size from the sources and the prompt. The snapshot text for the rest of that sentence is cut off, so read the page for what source returns.
| `size` value | No source images | With source images |
|---|---|---|
omitted or auto | Size picked from the prompt | Supported 2K size picked (default) |
source | Rejected | Accepted |
WIDTHxHEIGHT | Must be a 1K/2K preset | See the page |
What does size mean on Sume?
The request table says size is a shorthand for a resolution tier and that explicit pixels are not served in v1. The docs add: do not put custom pixels on size; use image_size or aspect_ratio. seed, output_compression and explicit pixel size are advertised by no model, so they return 400 unsupported_parameter. Tiers are 512, 1K, 2K and 4K.
How do I keep a reference's shape on Sume?
On edit and image-to-image calls, send aspect_ratio: "auto" to match the reference; the docs say omitting the field is not the same as auto. Combine it with a resolution tier your model lists. That gives you the shape of the input, though not a promise of the same pixel count: Sume's tiers are normalized, and exact pixels are a separate step.
If you need exact output pixels, resize afterward. See image API aspect ratios and custom sizes for what each ratio and tier gives.
Which one should I send?
On Ideogram: auto is the default with source images; the snapshot cuts off before saying what source returns, so read the page before relying on it. On Sume: aspect_ratio: "auto" for edits, or a named ratio for a fixed shape, and never a WIDTHxHEIGHT string in size. Read the model's supported_parameters before sending either.
Sources
Related posts
More in Developers
- Ideogram 4.5 edit images: 5 sources, first is edited, vs Sume
Ideogram 4.5 takes up to 5 source images, 25 MB each; the first is edited and multipart is required. Sume takes public HTTPS URLs as input_references.
- Ideogram 4 describe image to JSON prompt vs Sume reference input
Ideogram's describe endpoint returns a structured json_prompt with optional bounding boxes. The Sume docs list no such route; pass the image as a reference.
- Ideogram ad localizer API: exact_copy vs a Sume edit prompt
Ideogram's ad-localizer takes one language per call and an exact_copy mapping. On Sume you send the ad as a reference and spell out the copy in the prompt.
- Ideogram async generation_id polling vs Sume's 202 job envelope
Ideogram returns images directly unless async or webhook_url is set, then you poll /v2/generations. Sume blocks up to 30s, then returns a 202 job.
Written by Sume