Enhance Image Generation with Presets

Not yet available. Preset-driven image generation is tracked to be built — today, pass the parameters directly on POST /v1/images/generations (model, prompt, n, size), and manage presets in Dashboard → Settings → Presets.

Presets bundle a model and generation defaults under one name. Today they are managed in the dashboard; referencing a preset from an API call is part of the intended design. Image generation is billed per image in Algerian Dinars (DA).

Generate an image today#

The working call passes everything explicitly — model and prompt are required:

curl https://opendunes.com/api/v1/images/generations \
  -H "Authorization: Bearer $OPENDUNES_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-image-2",
    "prompt": "a dune at golden hour",
    "n": 1,
    "size": "1024x1024"
  }'

Intended design: reference a preset by id#

The intended design lets a request name a preset instead of repeating parameters. This does not work today — a request with a preset field but no model is rejected with a 422, because model is required:

# Intended design — not yet available
curl https://opendunes.com/api/v1/images/generations \
  -H "Authorization: Bearer $OPENDUNES_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "preset": "brand-hero", "prompt": "a dune at golden hour" }'

Why presets (intended design)#

Once preset-driven generation ships, presets are meant to give you:

• One source of truth for size, quality, and style across a team.
• The ability to swap the underlying model without touching client code.
• Per-preset budgets and rate limits.

Until then, keep the parameters in your client and manage preset definitions in Dashboard → Settings → Presets.

Next steps#

• Images API — the raw endpoint.
• Presets — manage presets programmatically.