# Z-Image Turbo (fal) - Core.Today AI API

> Tongyi-MAI's Z-Image Turbo, a 6B-parameter text-to-image model tuned for speed — a full image in as few as 8 steps. Priced per output megapixel, so smaller presets cost proportionally less, making it one of the cheapest options for high-volume generation.

- **Provider**: Tongyi-MAI
- **Model ID**: fal-ai/z-image/turbo
- **Category**: Image Generation
- **Credits**: 9 per image — 12 square_hd, 9 for 4:3 presets, 7 for 16:9 presets, 3 square
- **Speed**: Fast
- **Quality**: High

## Features
- Full image in as few as 8 inference steps
- Per-megapixel pricing — small presets cost proportionally less
- Six aspect-ratio presets from 512x512 to 1024x1024
- Up to 4 images per request
- Selectable acceleration level

## Use Cases
- High-volume thumbnail and variation generation
- Rapid prompt iteration before committing to a premium model
- Cost-sensitive batch pipelines
- Real-time or interactive image previews

## 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 - The prompt to generate an image from.

### Optional
- **image_size**: string (default: landscape_4_3) - Output size preset. Charged per output megapixel, so smaller presets cost less.
  Options: square_hd, square, portrait_4_3, portrait_16_9, landscape_4_3, landscape_16_9
- **num_inference_steps**: integer (default: 8) - The number of inference steps to perform (1-8).
- **num_images**: integer (default: 1) - Number of images to generate (1-4). Charged per image.
- **seed**: integer - The same seed and the same prompt will output the same image every time.
- **enable_safety_checker**: boolean (default: true) - If set to true, the safety checker will be enabled.
- **acceleration**: string (default: regular) - The acceleration level to use.
  Options: none, regular, high
- **output_format**: string (default: png) - The format of the generated image.
  Options: jpeg, png, webp

## Examples

### Cinematic portrait
Default landscape preset at 8 steps — the fastest, cheapest path to a usable image.
```json
{
  "model": "fal-ai/z-image/turbo",
  "input": {
    "prompt": "A hyper-realistic close-up portrait of a weathered lighthouse keeper, salt-crusted beard, storm light raking across his face, shot on a Leica M6 with Kodak Portra 400 grain",
    "image_size": "landscape_4_3",
    "num_inference_steps": 8,
    "num_images": 1,
    "output_format": "png"
  }
}
```

### Low-cost square batch
The 512x512 square preset is a quarter of a megapixel, so a 4-image batch costs about the same as one square_hd image.
```json
{
  "model": "fal-ai/z-image/turbo",
  "input": {
    "prompt": "Flat vector icon of a paper airplane, bold single-color silhouette, generous margins, centered",
    "image_size": "square",
    "num_images": 4,
    "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
- Pricing follows output megapixels: 'square' (512x512) costs a quarter of 'square_hd' (1024x1024). Pick the smallest preset that fits your use.
- This is a turbo model — num_inference_steps caps at 8, and raising it past the default rarely improves the result.
- Custom width/height is not exposed here; use one of the six presets. For arbitrary dimensions, use prunaai/z-image-turbo on Replicate instead.
- Distinct from prunaai/z-image-turbo: that is Pruna's optimised build with width/height inputs, this is the original at half the cost per megapixel.

## Documentation
https://fal.ai/models/fal-ai/z-image/turbo