Images

Use it when

• Your app already uses the OpenAI Images API shape.
• You want to generate one or more images from a prompt with the same Prismatic API key as text endpoints.
• The selected model and provider route both support images.

Code examples

curl https://api.prismaticapi.com/v1/images/generations   -H "Authorization: Bearer $PRISMATIC_API_KEY"   -H "Content-Type: application/json"   -d '{
    "model": "imagen-4",
    "prompt": "A clean product mockup on a neutral desk",
    "image_url": ["https://example.com/reference.png"],
    "negative_prompt": "clutter, text artifacts",
    "aspect_ratio": "16:9",
    "quality": "high",
    "style": "natural",
    "seed": 42,
    "response_format": "b64_json",
    "n": 1
  }'

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.PRISMATIC_API_KEY,
  baseURL: "https://api.prismaticapi.com/v1"
});

const image = await client.images.generate({
  model: "imagen-4",
  prompt: "A clean product mockup on a neutral desk",
  image_url: "https://example.com/reference.png",
  negative_prompt: "clutter, text artifacts",
  aspect_ratio: "16:9",
  quality: "high",
  style: "natural",
  seed: 42,
  response_format: "b64_json",
  n: 1
});

Parameters

• model (string, required) — Public model ID returned by GET /v1/models.
• prompt (string, required) — Text prompt describing the image to generate.
• image_url (string or array, optional) — Reference image URL(s) forwarded natively or converted to multimodal chat content for chat-completions providers.
• negative_prompt, aspect_ratio, seed, quality, style, response_format, size and n are accepted and forwarded when supported by the selected provider route.

Billing and limits

• API-key and account rate limits are checked before generation.
• Subscription quota is reserved before generation and settled afterward.
• PAYG generation cost is retail_generation_usd * n; requests are refused before upstream when generation pricing is missing or the wallet cannot cover the precise estimate.
• If the model or provider route does not support images, the request fails before upstream calls.

Notes

• GET /v1/models can list multiple public IDs that resolve to the same canonical model.
• Mark a provider with the Images toggle in admin so it can receive image-generation traffic.
• The endpoint rewrites the model to the upstream ID before proxying.