AI tattoo generator: design ideas and a preview on your skin
Ask an AI for tattoo designs as black linework on white, preview one on a photo of your arm by sending both images, then take it to a tattoo artist.

To use an AI tattoo generator, ask for the design as clean black linework on a plain white background, the look of a flash sheet, and compare a few versions. To see a design on your body, send a photo of the spot and the design to an image model together and describe the placement and size. Then take the design to a tattoo artist to redraw for your skin. On Sume, ChatGPT Image 2.5 today returns up to 4 designs per call and takes up to 16 reference images, so the photo and the design fit in one request.
Sume facts come from the Image API docs and the request schemas in the Sume API reference, read on 2026-09-28.
How do I write a tattoo design prompt?
Describe the design the way you would to an artist, and send it to POST /v1/images with openai/gpt-image-2.5:
- The subject and style: "a fine-line peony with two leaves", "a traditional swallow with bold outlines", or "a geometric wolf head".
- The flash-sheet look: "black linework on a plain white background, no shading", or "with dotwork shading".
- The shape of the spot: a long vertical design for a forearm, a small round one for a wrist. Match
aspect_ratioto it, such as9:16or1:1. nfor up to 4 versions per call. Each completed image is billed.
curl -X POST "https://api.sume.com/v1/images" \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "openai/gpt-image-2.5",
"prompt": "Tattoo design: a fine-line peony with two leaves, drawn as a long vertical piece for a forearm. Black linework on a plain white background, no shading, no text.",
"aspect_ratio": "9:16",
"n": 4
}'How do I see a tattoo design on my body?
Take a well-lit photo of the spot, put it and the design at public HTTPS URLs, and send both in input_references. It is the same two-image request as an AI mockup: the prompt names each image by what it shows, and aspect_ratio is set to your photo's shape, such as 3:4, rather than "auto". Then give the placement, the size, and what must stay.
{
"model": "openai/gpt-image-2.5",
"prompt": "Place the peony from the line drawing on the inner forearm in the arm photo, about a hand's length tall, as black ink following the curve of the arm. Keep the arm, the skin, and the background as they are.",
"input_references": [
{ "type": "image_url", "image_url": { "url": "https://example.com/forearm.jpg" } },
{ "type": "image_url", "image_url": { "url": "https://example.com/peony-lines.png" } }
],
"aspect_ratio": "3:4",
"n": 4
}Can I turn a photo into a tattoo design?
Yes. Send the photo, such as a pet, a flower, or a skyline, as the only reference and ask for "black linework on white, simplified for a tattoo". With one reference, aspect_ratio: "auto" keeps the photo's shape.
To try the design on other photos in your own editor, cut out the white background: POST /v1/rmbg-1.0/remove returns a PNG with alpha. API pricing lists Remove background at $0.0225 per image, plus a 5.5% agent fee by default, and remove background API covers the call.
Which fields does each step use?
Both steps are the same POST /v1/images request with different inputs. Only models that edit from references accept input_references; a text-to-image-only model rejects them.
| Field | Design sheet | Preview on your body |
|---|---|---|
model | openai/gpt-image-2.5 | openai/gpt-image-2.5 |
prompt | Subject, style, and "black linework on white" | Which image is which, the placement, the size, and what stays |
input_references | None, or one photo to turn into line art | The body photo and the design (this model takes up to 16) |
aspect_ratio | The design's shape, such as 9:16 or 1:1 | Your photo's shape, such as 3:4 |
n | Up to 4 | Up to 4 |
What are the limits?
- The preview is a redrawn photo, not a simulation of ink in skin. The model repaints the whole image, so your arm, your skin, or an existing tattoo can change too.
- An artist decides what works on skin: line weight, size, and detail. Bring the AI design as a reference, not as a stencil.
- Today designs come back as PNG, JPEG, or WebP. No model lists
svg, so there is no vector file. - Each completed image is billed, and a failed one is not. ChatGPT Image 2.5 estimates each request's price from its size and quality, and
usage.costin the response is the USD amount billed. - Results are Sume-hosted, signed URLs, so download the designs you keep.
Sources
Related posts
More in Use cases
- AI training video generator: from written steps to video
An AI training video generator turns written steps into presenter videos. Build one module from short avatar clips: one clip per step, captioned, joined.
- Arcads alternatives: AI UGC ad tools by plan and API
Arcads alternatives for AI-actor UGC ads: Creatify, HeyGen, Higgsfield and Sume compared on presenters, API, billing and free options, read 2026-09-28.
- Arcads vs Creatify: AI actor ads or URL-to-video ads
Arcads builds ads from your script and 1,000+ AI actors on monthly plan credits. Creatify turns a product URL into a video ad; API plans start at $99.
- Can you schedule YouTube Shorts? Yes, in Studio or by API
Yes. Upload a Short in YouTube Studio and set a publish time on the Schedule card, or set status.publishAt on a private upload with the Data API.
Written by Sume