# Moondream2 - Core.Today AI API

> Small but capable vision-language model (14M+ runs) — ask free-form questions about any image and get detailed answers. Efficient VQA for tagging, moderation prep, and rich alt-text.

- **Provider**: Moondream
- **Model ID**: lucataco/moondream2
- **Category**: Image Generation
- **Credits**: 5 per image
- **Speed**: Fast
- **Quality**: High

## Features
- Free-form questions about any image
- Detailed multi-sentence answers
- Small model — fast and cheap (7 credits)
- Good at counting, reading, and describing

## Use Cases
- Rich image descriptions
- Content moderation pre-screening
- Product attribute extraction
- Chart and diagram explanation
- Accessibility alt-text

## 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
- **image**: string - Input image

### Optional
- **prompt**: string (default: Describe this image) - Question or instruction about the image

## Examples

### Detailed Description
Get a rich description of an image
```json
{
  "model": "lucataco/moondream2",
  "input": {
    "image": "https://example.com/scene.jpg",
    "prompt": "Describe this image in detail, including colors, objects, and mood."
  }
}
```

### Product Attributes
Extract structured attributes from a product photo
```json
{
  "model": "lucataco/moondream2",
  "input": {
    "image": "https://example.com/product.jpg",
    "prompt": "List the product's color, material, and any visible text."
  }
}
```

## 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
- Ask specific questions — 'What brand is the logo?' beats 'Describe this'
- Great mid-point between 1-credit BLIP captions and expensive frontier VLMs
- Answers are in English; translate downstream if needed

## Documentation
https://replicate.com/lucataco/moondream2