Skip to content
Poremu
Back to app

Developer docs

Generate images

Text-to-image and image-to-image requests with full parameter reference.

text
POST /v1/generations

Request

json
{
  "model": "gpt-image",
  "prompt": "Editorial photo of a glass perfume bottle on wet black stone, rim light",
  "aspectRatio": "3:4",
  "quality": "high",
  "count": 2,
  "idempotencyKey": "order-8812-hero"
}
FieldTypeNotes
modelstringRequired. A model id from /v1/models.
promptstringRequired. Up to 4 000 characters.
negativePromptstringOnly for models with capabilities.negativePrompt.
inputImagesstring[]Public https URLs (or URLs returned by uploads). Enables edit / reference mode.
aspectRatiostringOne of the model's aspectRatios.
resolution"1k" | "2k" | "4k"Only for models that list resolutions.
quality"low" | "medium" | "high"Only for models that list qualities.
seedintegerOnly for models with capabilities.seed.
count1–4Number of images. Each is billed.
idempotencyKeystringSame key within your account returns the same generation instead of creating a new one. Recommended for retries.
webhookUrlstringhttps URL called once the job finishes. See Webhooks.

Response 201

json
{
  "id": "6f1d6a2e-7f2d-4bd9-9e2e-0f3a1b2c3d4e",
  "status": "QUEUED",
  "model": "gpt-image",
  "modality": "image",
  "prompt": "Editorial photo of …",
  "params": { "aspectRatio": "3:4", "quality": "high", "count": 2 },
  "inputImages": [],
  "outputs": [],
  "credits": 120,
  "source": "API",
  "createdAt": "2026-03-02T04:10:11.000Z"
}

credits is the amount reserved. When the job succeeds, outputs[] is filled:

json
"outputs": [
  {
    "id": "a1…",
    "kind": "IMAGE",
    "url": "https://cdn.poremu.ai/u/…/0.png",
    "thumbnailUrl": "https://cdn.poremu.ai/u/…/0.thumb.webp",
    "width": 1024,
    "height": 1536,
    "mimeType": "image/png"
  }
]

Image-to-image

Pass one or more inputImages. The model decides how to use them: gpt-image and qwen-image-3 treat them as the image to edit, nano-banana uses them as references/characters.

bash
curl https://api.poremu.ai/v1/generations \
  -H "Authorization: Bearer $POREMU_API_KEY" -H "Content-Type: application/json" \
  -d '{
    "model": "nano-banana",
    "prompt": "Same person, now wearing a navy wool coat, walking in Seoul at dusk",
    "inputImages": ["https://cdn.poremu.ai/uploads/…/ref.jpg"],
    "aspectRatio": "9:16"
  }'

JavaScript

js
const res = await fetch('https://api.poremu.ai/v1/generations', {
  method: 'POST',
  headers: { Authorization: `Bearer ${process.env.POREMU_API_KEY}`, 'Content-Type': 'application/json' },
  body: JSON.stringify({ model: 'nano-banana', prompt: 'A cosy reading nook, watercolor', aspectRatio: '4:3' }),
});
const gen = await res.json();
const done = await fetch(`https://api.poremu.ai/v1/generations/${gen.id}/wait?timeout=60000`, {
  headers: { Authorization: `Bearer ${process.env.POREMU_API_KEY}` },
}).then((r) => r.json());
console.log(done.outputs.map((o) => o.url));

Python

python
import os, requests

H = {"Authorization": f"Bearer {os.environ['POREMU_API_KEY']}"}
gen = requests.post("https://api.poremu.ai/v1/generations", headers=H,
                    json={"model": "gpt-image", "prompt": "Isometric icon of a rocket, pastel", "quality": "medium"}).json()
done = requests.get(f"https://api.poremu.ai/v1/generations/{gen['id']}/wait", headers=H, params={"timeout": 60000}).json()
print([o["url"] for o in done["outputs"]])