Recraft V4 or FLUX.2 Pro for an illustration set on Sume: price, edits
Recraft V4 bills $0.05 per image on Sume and is text-only; FLUX.2 Pro bills $0.0375 and takes reference images. Ratios, n limits and which fits a set.

FLUX.2 Pro is the cheaper of the two on Sume and the only one that takes reference images: $0.0375 billed per image against $0.05 for Recraft V4, and reference images against none. Recraft V4 is text-to-image only. For an illustration set that must hold one look across a dozen images, the practical difference is that FLUX.2 Pro can be shown the first approved image and asked for the next, while Recraft V4 can only be steered by the prompt.
Everything else on the request side is the same, which makes a swap easy. Both list the same 13 aspect ratios and both take up to 4 images per call.
The side-by-side facts
The figures below come from Sume's image catalog and price tables. Billed is provider list times 1.25.
| Field | FLUX.2 Pro | Recraft V4 |
|---|---|---|
| Model id | black-forest-labs/flux.2-pro | recraft/recraft-v4 |
| List per image | $0.03 | $0.04 |
| Billed per image | $0.0375 | $0.05 |
| Reference images | Yes, image edits | None, text-only |
| Images per call (n) | Up to 4 | Up to 4 |
| Aspect ratios | 13, from 1:2 to 21:9 | 13, from 1:2 to 21:9 |
| Custom pixels via image_size | Yes | Yes |
| resolution and quality tiers | None | None |
| 12-image set, billed | $0.45 | $0.60 |
Why references decide a set
A set is only a set if the images agree on palette, line weight and subject style. With no reference input, agreement has to come from repeating the same style words in each prompt. With FLUX.2 Pro you can attach the approved first image through input_references and ask for the next scene in the same style.
That is a mechanism, not a guarantee. Run a small test: generate a first image, then three follow-ups with and without the reference, and compare. The whole test is four FLUX.2 Pro images with a reference ($0.15) plus three Recraft images without ($0.15).
What the numbers say about a 12-image set
Twelve images are $0.45 billed on FLUX.2 Pro and $0.60 on Recraft V4. Even if every image needed two takes, FLUX.2 Pro would cost $0.90 and Recraft $1.20. The money is not the point at this size; the choice should turn on the reference input and on how the images look to you. Run the same twelve prompts through both and see.
At catalogue scale the gap becomes real. A thousand images are $37.50 on FLUX.2 Pro and $50.00 on Recraft V4, a $12.50 difference, or 33 percent more for Recraft.
Ratios an illustration set usually needs
Both models list 1:1, 16:9, 9:16, 4:3, 3:4, 4:5, 5:4, 3:2, 2:3, 21:9, 9:21, 1:2 and 2:1. That covers a square avatar set, a 16:9 hero, a 9:16 story frame and the tall 1:2 and wide 2:1 strips. If you need an exact pixel size that the ratios do not give you, both accept image_size with a width and height. Sume's catalog returns 400 unsupported_parameter for any field a model does not list, so a ratio typo fails fast and is not billed.
Neither model has a quality or resolution field, so there is no tier to tune. The only cost levers are the model choice and n. Because n multiplies the per-image price, n: 4 on Recraft V4 costs $0.20 billed and on FLUX.2 Pro $0.15.
Choosing in one line
If a set of images has to look consistent and your starting point is an existing photo or illustration, use FLUX.2 Pro. If you are starting from a blank prompt and want the second-cheapest non-draft card that supports 21:9 and 2:1, Recraft V4 is a reasonable test. Compare the two on your own prompts before committing, since this post compares only what Sume lists and bills.
Request shape
The call below is identical for both models except for the model value and the optional reference. If you send input_references to Recraft V4, Sume returns 400 unsupported_parameter because the model lists no reference range.
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": "Flat illustration of a greenhouse at dusk, muted teal and ochre",
"aspect_ratio": "3:2",
"n": 2
}'Sources
Related posts
More in Comparisons
- Runway Dev usage exception request vs Sume 429 and 402
Runway Dev says higher usage needs an exception request. On Sume a burst hits 429 or 402. What each means for a batch of ad variants and how to retry.
- Runway Dev image_to_video, video_to_video: one Sume route?
Runway Dev names image_to_video, video_to_video and text_to_image paths. Sume's POST /v1/videos infers the mode from your fields. What it changes in ad code.
- Seedream 5.0 Lite or Nano Banana 2.1 for product shots on Sume
Seedream 5.0 Lite bills $0.04375 per image on Sume against $0.10 for Nano Banana 2.1 at 1K. Ratios, tiers, references and when the extra cost is worth paying.
- Text-to-video or image-to-video: when the prompt alone is enough
Use text-to-video when the look is open, and image-to-video when a frame, a face or a product must match. The Sume models that take each, and a decision rule.
Written by Sume