# ControlNet Scribble - Core.Today AI API

> The classic sketch-to-image model (38M+ runs) — turn any scribble or line drawing into a detailed image guided by your prompt. Draw the composition, describe the content.

- **Provider**: ControlNet
- **Model ID**: jagilley/controlnet-scribble
- **Category**: Image Generation
- **Credits**: 120 per run (num_samples '4' costs 480)
- **Speed**: Medium
- **Quality**: High

## Features
- Sketch controls composition, prompt controls content
- 38M+ runs — the scribble standard
- 1 or 4 samples per run
- Additional prompt / negative prompt refinement

## Use Cases
- Concept art from rough sketches
- Whiteboard drawing to illustration
- Kid-drawing gift art
- Layout-controlled product shots
- Storyboard visualization

## 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 - What the final image should be
- **image**: string - Scribble/line drawing input

### Optional
- **num_samples**: string (default: 1) - Number of images ('1' or '4'). '4' is billed 4x
  Options: 1, 4
- **image_resolution**: string (default: 512) - Output resolution
  Options: 256, 512, 768
- **ddim_steps**: integer (default: 20) - Denoising steps
- **scale**: number (default: 9) - Prompt guidance scale (0.1-30)
- **seed**: integer - Random seed
- **eta**: number (default: 0) - eta (DDIM)
- **a_prompt**: string (default: best quality, extremely detailed) - Added quality prompt
- **n_prompt**: string (default: longbody, lowres, bad anatomy, bad hands, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality) - Negative prompt

## Examples

### Sketch to Illustration
Turn a rough house sketch into a detailed illustration
```json
{
  "model": "jagilley/controlnet-scribble",
  "input": {
    "image": "https://example.com/house-sketch.png",
    "prompt": "a cozy cottage in a forest, warm sunset light, storybook illustration",
    "image_resolution": "512"
  }
}
```

## 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
- Bold, clear lines in the sketch give the strongest composition control
- num_samples '4' explores variations quickly (billed 4x)
- For sketch editing with frontier quality, also try nano-banana or FLUX Kontext with the sketch attached

## Documentation
https://replicate.com/jagilley/controlnet-scribble