FLUX 3 Image

FLUX 3 Image lets you build a picture out of several others: the lamp from one photo in the room from another, or three products arranged on a poster in a fixed spot. It is the still-image model of Black Forest Labs' FLUX 3 family, next to FLUX 3 Video. Each request makes or edits one image from a prompt and up to ten references.

You can also mark where things go with boxes written into the prompt, which helps with posters and mockups.

What you can make

• Object swaps, for example "the sofa from ref_image_1 in the living room from ref_image_0"
• Posters and mockups where the logo, product and headline sit in set positions
• Cheap 768sq previews before committing to a 2K render
• Group scenes made from products or characters photographed separately

Settings that matter

• image_urls: up to 10. The prompt refers to them as ref_image_0, ref_image_1 and so on, so keep the order fixed.
• resolution: 768sq, 1k (default), 1.5k, 2k or 4k. 4K costs several times more than 2K; use it for the final.
• aspect_ratio: 15 formats from 21:9 to 9:21, or auto.
• Layout: add a small table with id, bbox (0-1000 coordinates) and desc. Boxes guide placement; they are not masks.
• When editing, name what must not change, such as the floor, window or camera angle.

Starter prompt

Put the armchair from ref_image_1 next to the window in
ref_image_0. Keep the floor, wall color, window and
camera angle unchanged. Match the room's soft daylight.

Swap the object and its position; list everything else that has to stay as it is.

Questions

What does an image cost?

It depends on the resolution tier: from $0.0984 per image. The full table is on this page and in pricing. A failed image is free.

Which settings does it not have?

There is no seed, no pixel dimensions, no steps, no guidance and no negative prompt. You steer it with the prompt, resolution, ratio and references.

When is FLUX.2 [pro] enough?

For simple 1K or 2K graphics, FLUX.2 [pro] costs less. Pick FLUX 3 Image when you need more than 8 references, a 4K tier or layout boxes.

FLUX 3 Image in your code

curl https://kadraris.com/v1/generate?wait=60 \
  -H "Authorization: Bearer $KADRARIS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "flux-3-image",
  "input": {
    "prompt": "Packshot of a white linen sneaker on a concrete block, hard midday sun, crisp shadow, 85mm",
    "aspect_ratio": "auto",
    "resolution": "1k"
  }
}'

import os, time, requests

API = "https://kadraris.com/v1"
H = {"Authorization": f"Bearer {os.environ['KADRARIS_API_KEY']}"}

task = requests.post(f"{API}/generate", headers=H, json={
    "model": "flux-3-image",
    "input": {
        "prompt": "Packshot of a white linen sneaker on a concrete block, hard midday sun, crisp shadow, 85mm",
        "aspect_ratio": "auto",
        "resolution": "1k"
    }
}).json()
while task["status"] in ("queued", "running"):
    time.sleep(3)
    task = requests.get(f"{API}/tasks/{task['id']}", headers=H).json()

print(task["status"], task["cost"], [o["url"] for o in task["output"]])

const API = "https://kadraris.com/v1";
const headers = { Authorization: `Bearer ${process.env.KADRARIS_API_KEY}`, "Content-Type": "application/json" };

let task = await (await fetch(`${API}/generate`, {
  method: "POST",
  headers,
  body: JSON.stringify({
    "model": "flux-3-image",
    "input": {
      "prompt": "Packshot of a white linen sneaker on a concrete block, hard midday sun, crisp shadow, 85mm",
      "aspect_ratio": "auto",
      "resolution": "1k"
    }
  }),
})).json();

while (task.status === "queued" || task.status === "running") {
  await new Promise((r) => setTimeout(r, 3000));
  task = await (await fetch(`${API}/tasks/${task.id}`, { headers })).json();
}
console.log(task.status, task.cost, task.output.map((o) => o.url));

Input fields

Fetch this list as JSON with GET /v1/models/flux-3-image.