FLUX 3 Image API

FLUX 3 Image is the image model of Black Forest Labs' FLUX 3 family (the video model is FLUX 3 Video). It generates or edits one image per request from a prompt and up to 10 reference images. Reference order matters: in the prompt you call them ref_image_0, ref_image_1 and so on. Resolution comes in five tiers: 768sq, 1K, 1.5K, 2K and 4K, and the aspect ratio is one of 15 formats from 21:9 to 9:21, or auto. You can guide where elements go with boxes (0-1000 coordinates) written into the prompt text.

Best for

• Swapping an object in a photo: "the chair from ref_image_1 in the room from ref_image_0"
• Posters and mockups where text and product must sit in a set place
• Quick 768sq previews before a 2K render
• Scenes assembled from several products or characters shot separately

Prompting tips

• When editing, say plainly what stays the same: the window, the floor, the camera angle.
• Keep the reference order stable, the model knows them by number.
• Put the layout in the prompt as a table with id, bbox and desc - boxes guide placement, they are not masks.
• 4K costs several times more than 2K, so save it for the final version.

FAQ

How much does FLUX 3 Image cost?

The price depends on resolution: from $0.0984 per image. The full tier table is on this page and in pricing. Failed generations are free.

Does FLUX 3 Image support seed or a negative prompt?

No. This model does not accept seed, pixel dimensions, steps, guidance or a negative prompt. You steer it with the prompt, resolution, aspect ratio and references.

FLUX 3 Image or FLUX.2 [pro]?

FLUX.2 [pro] is cheaper at 1K and 2K. FLUX 3 Image takes more references (10 instead of 8), has a 4K tier and supports layout boxes in the prompt. For simple 1K graphics FLUX.2 is enough.

How to call FLUX 3 Image via API

curl https://jantaris.com/v1/generate?wait=60 \
  -H "Authorization: Bearer $JANTARIS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "flux-3-image",
  "input": {
    "prompt": "Product photo of a matte black ceramic coffee cup on a travertine table, soft window light, 50mm",
    "aspect_ratio": "auto",
    "resolution": "1k"
  }
}'

import os, time, requests

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

task = requests.post(f"{API}/generate", headers=H, json={
    "model": "flux-3-image",
    "input": {
        "prompt": "Product photo of a matte black ceramic coffee cup on a travertine table, soft window light, 50mm",
        "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://jantaris.com/v1";
const headers = { Authorization: `Bearer ${process.env.JANTARIS_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": "Product photo of a matte black ceramic coffee cup on a travertine table, soft window light, 50mm",
      "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));

Parameters

The same schema is available from GET /v1/models/flux-3-image.