Images

POST https://dev.typno.ai/v1/images/generations makes an image from a description.

curl https://dev.typno.ai/v1/images/generations \
  -H "Authorization: Bearer $TYPNO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "z-image-turbo", "prompt": "a lighthouse at sunset, flat illustration", "size": "1024x1024"}'

Field
`model`	An image model, see below
`prompt`	What the image shows; English works best
`size`	`WIDTHxHEIGHT`, 256 to 2048 pixels per side. Default `1024x1024`
`n`	Number of images, 1 to 4. Default 1

The answer has the images as base64 in data[].b64_json.

import base64

result = client.images.generate(model="z-image-turbo", prompt="a lighthouse at sunset")
open("lighthouse.jpg", "wb").write(base64.b64decode(result.data[0].b64_json))

Models

Model	Per megapixel
Flux.2 [klein] `flux2-klein`	€ 0,0055
Z-Image Turbo `z-image-turbo`	€ 0,0099
Qwen Image `qwen-image`	€ 0,0825

Images are billed per megapixel of output.