Image Generation - NagaAI Documentation

POST /v1/images/generations creates images from a prompt. Use this endpoint when your workflow is primarily about generating images, not running a broader multimodal conversation.

Common fields

Field	Common use
`model`	image generation model
`prompt`	image description or instruction
`size`	output dimensions
`response_format`	`url` or `b64_json`
`n`	number of images, when supported by the model

Example Request

from openai import OpenAI

client = OpenAI(
    base_url="https://api.naga.ac/v1",
    api_key="YOUR_API_KEY",
)

result = client.images.generate(
    model="gpt-image-1",
    prompt="A cinematic watercolor fox reading in a lantern-lit library.",
    size="1024x1024",
    response_format="url",
)

print(result.data[0].url)

import OpenAI from 'openai';

const client = new OpenAI({
  baseURL: 'https://api.naga.ac/v1',
  apiKey: 'YOUR_API_KEY',
});

const result = await client.images.generate({
  model: 'gpt-image-1',
  prompt: 'A cinematic watercolor fox reading in a lantern-lit library.',
  size: '1024x1024',
  response_format: 'url',
});

console.log(result.data[0].url);

curl https://api.naga.ac/v1/images/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-image-1",
    "prompt": "A cinematic watercolor fox reading in a lantern-lit library.",
    "size": "1024x1024",
    "response_format": "url"
  }'

Response Shape

{
  "created": 1710000000,
  "data": [
    {
      "url": "https://api.naga.ac/outputs/..."
    }
  ],
  "usage": {
    "input_tokens": 42,
    "output_tokens": 512,
    "total_tokens": 554
  }
}

Notes

response_format can be url or b64_json
n and size constraints depend on the model
the prompt is moderated before generation

Common mistakes

using Images API for workflows that should really be multimodal Responses prompts
assuming every model supports the same size or n combinations
forgetting to handle b64_json differently from hosted image URLs

Reference

Generate images