Web Analytics Made Easy - Statcounter
Skip to content

Quick Start

Get started with Duckling in 5 minutes.

Starting the Application

Choose your preferred method:

The fastest way to get started - no dependencies to install!

Option 1: Pre-built Images (Fastest) 

# Download the compose file
curl -O https://raw.githubusercontent.com/davidgs/duckling/main/docker-compose.prebuilt.yml

# Start Duckling
docker-compose -f docker-compose.prebuilt.yml up -d

Option 2: Build Locally 

# Clone and start
git clone https://github.com/davidgs/duckling.git
cd duckling
docker-compose up --build

The UI will be available at http://localhost:3000

First Run

The first startup may take a few minutes as Docker downloads/builds the images.

Terminal 1: Backend

cd backend
source venv/bin/activate  # Windows: venv\Scripts\activate
python duckling.py

The API will be available at http://localhost:5001

Terminal 2: Frontend

cd frontend
npm run dev

The UI will be available at http://localhost:3000

Your First Conversion

1. Open the Application

Navigate to http://localhost:3000 in your browser.

Duckling Interface
The main Duckling interface

2. Upload a Document

Drag and drop a PDF, Word document, or image onto the drop zone, or click to browse.

3. Watch the Progress

The conversion progress will be displayed in real-time.

4. Download Results

Once complete, choose your export format:

Conversion Complete
Conversion complete with export options
  • Markdown - Great for documentation
  • HTML - Web-ready output
  • JSON - Full document structure
  • Plain Text - Simple text extraction

Basic Configuration

Click the Settings button to configure:

OCR Settings

Setting Default Description
Enabled true Enable OCR for scanned documents
Backend easyocr OCR engine to use
Language en Primary language

Table Settings

Setting Default Description
Enabled true Extract tables from documents
Mode accurate Detection accuracy level

Image Settings

Setting Default Description
Extract true Extract embedded images
Scale 1.0 Image output scale

Batch Processing

To convert multiple files at once:

  1. Drag and drop multiple files or a whole folder onto the drop zone. The browser expands a folder into its files; Duckling queues every supported document (unsupported types are skipped).
  2. Click the drop zone to open a folder picker and upload all supported files inside that folder at once.
  3. Use Choose files… when you want to pick individual files only (not folder mode).

All queued files are processed according to the job queue (see Features for concurrent limits).

Performance

Batch processing uses a job queue with a maximum of 2 concurrent conversions to prevent memory exhaustion.

Using the API

For programmatic access, use the REST API:

# Upload and convert a document
curl -X POST http://localhost:5001/api/convert \
  -F "file=@document.pdf"

# Response
{
  "job_id": "550e8400-e29b-41d4-a716-446655440000",
  "status": "processing"
}

Check the API Reference for complete documentation.

Next Steps