Image generation

POST /api/v1/images/generations generates an image from a text prompt. Compatible with the OpenAI Images API.

Request

curl https://api.pendra.ai/api/v1/images/generations \
  -H "Authorization: Bearer pdr_sk_..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "x/z-image-turbo",
    "prompt": "A misty Welsh valley at dawn, painted in oils",
    "size": "1024x1024"
  }'

Fields

• model — image model ID, e.g. x/z-image-turbo. List models at /models?type=image.
• prompt — the text description.
• size — e.g. 1024x1024. Supported sizes depend on the model.
• n — number of images to generate (model permitting; default 1).
• response_format — b64_json (default) or url depending on the underlying backend.

Response

OpenAI-shaped image response. data contains one entry per requested image (set by n). With the default response_format: "b64_json", each entry carries a base64-encoded PNG — decode it before writing to disk. Backends that return URLs (some hosted models, set response_format: "url") populate url instead.

{
  "created": 1715346200,
  "data": [
    {
      "b64_json": "iVBORw0KGgoAAAANSUhEUgAA...Cg=="
    }
  ]
}

Python

from pendra import Pendra
import base64, pathlib

client = Pendra()
result = client.images.generate(
    model="x/z-image-turbo",
    prompt="A misty Welsh valley at dawn",
    size="1024x1024",
)

pathlib.Path("out.png").write_bytes(base64.b64decode(result.data[0].b64_json))

Timeouts

Image requests have a longer ceiling than chat — up to ~600 seconds — because generation can be slow on busy workers. For interactive UX, render a spinner; you'll usually get a response in 5–30 seconds depending on model and size.

Usage tracking

Image requests appear under Images in the console usage view alongside chat and embeddings.