# GPT Image 2 - Core.Today AI API

> OpenAI's state-of-the-art image generation and editing model with strong instruction following, sharp text rendering, and detailed editing. Quality-based pricing lets you trade off cost vs. fidelity.

- **Provider**: OpenAI
- **Model ID**: openai/gpt-image-2
- **Category**: Image Generation
- **Credits**: 300 per image (high/auto); 110 for medium, 29 for low
- **Speed**: Medium
- **Quality**: Ultra

## Features
- Strong prompt and instruction following
- Sharp typography and text rendering
- Image editing via input_images (composition/edit)
- Quality tiers: low / medium / high / auto
- Transparent or opaque background control

## Use Cases
- Marketing visuals with legible copy
- Photoreal portraits and product shots
- Multi-image composition and edits
- Concept art with strong text elements

## API Endpoint

Base URL: https://api.core.today/v1
Create Prediction: POST /predictions
Get Status: GET /predictions/{job_id}

## Authentication
Header: X-API-Key: YOUR_API_KEY

## Input Parameters

### Required
- **prompt**: string - A text description of the desired image

### Optional
- **openai_api_key**: string - Your OpenAI API key (optional - uses proxy if not provided)
- **aspect_ratio**: string (default: 1:1) - Output aspect ratio
  Options: 1:1, 3:2, 2:3, 4:3, 3:4, 16:9, 9:16, auto, 1024x1024, 1536x1024, 1024x1536, 1536x1152, 1152x1536, 2048x2048, 2048x1152, 1152x2048, 3840x2160, 2160x3840
- **input_images**: array - Array of image URLs to use as input for editing or composition
- **number_of_images**: integer (default: 1) - Limited to 1 — call multiple times for batches
- **quality**: string (default: auto) - Quality tier — sharper but slower and more expensive at higher tiers
  Options: low, medium, high, auto
- **background**: string (default: auto) - Set background mode (transparent backgrounds not supported)
  Options: auto, transparent, opaque
- **output_compression**: integer (default: 90) - Compression level (0-100) for webp/jpeg output
- **output_format**: string (default: webp) - Output image format
  Options: png, jpeg, webp
- **moderation**: string (default: auto) - Content moderation level
  Options: auto, low
- **user_id**: string - An optional unique identifier representing your end-user. This helps OpenAI monitor and detect abuse.

## Examples

### Photorealistic Portrait
Generate a high-fidelity portrait with rich detail
```json
{
  "model": "openai/gpt-image-2",
  "input": {
    "prompt": "A photorealistic close-up portrait of an elderly Japanese ceramicist with deep wrinkles and a warm smile, carefully inspecting a freshly glazed tea bowl in his rustic sun-drenched workshop.",
    "aspect_ratio": "3:2",
    "quality": "high",
    "output_format": "webp"
  }
}
```

## Response Format
```json
{
  "job_id": "abc123",
  "status": "pending | processing | completed | failed",
  "result": "URL or data (when completed)"
}
```

## Usage Flow
1. POST /predictions with model and input -> receive job_id
2. GET /predictions/{job_id} -> poll until status is completed or failed
3. Result contains output URL(s)

## Tips
- Use quality=low for cheap iterations, then re-render at high once happy
- Provide input_images to compose or edit existing visuals
- GPT Image 2 excels at images that require legible text
- Use 3:2 / 2:3 for cinematic or portrait orientation