cicora.ai
English
API specificationImage generation

Images

Create images through one route and select parameters from the capabilities of the chosen model.

Integration reference: choose parameters and available capabilities from your account API settings and the model catalogue for your environment.

Create a generation

A request contains model, prompt, and only applicable artifact parameters: size, aspect ratio, format, or result count.

Do not use a text model as an image identifier: select the model from its catalogue modality.

Image generation request
bash
curl --request POST https://cicora.ai/api/v1/images \
  --header "Authorization: Bearer $CICORA_API_KEY" \
  --header 'Content-Type: application/json' \
  --data '{
    "model": "google/gemini-2.5-flash-image",
    "prompt": "A precise editorial illustration of a delivery route at dusk",
    "aspect_ratio": "1:1",
    "resolution": "1K",
    "n": 1
  }'

References and editing

  • Send reference images through input_references only in the form supported by the selected model.
  • Image-to-image and editing use the same /images route with the source reference and a textual instruction.
  • Check mime type, size, and rights for an input file before sending it.

Receive the artifact

In the compatible response, read data[].b64_json as image bytes; media_type can identify the format. Retain metadata with its prompt and model.

Do not publish a partial result as final. With streaming delivery, distinguish a draft event from a completed artifact.

Continue with