Skip to content

Quick Start

Extract text from your first document in under 5 minutes.

1. Start the Stack

just dev
# Or: docker compose -f docker/docker-compose.yml up -d

Wait for all services to be healthy:

just status

2. Open the API Docs

Visit http://localhost:8000/api/docs for the interactive Swagger UI.

3. Extract Text from a PDF

curl -X POST http://localhost:8000/api/v1/extract \
  -F "file=@my-document.pdf" \
  -F "output_format=markdown"

import httpx

with open("my-document.pdf", "rb") as f:
    response = httpx.post(
        "http://localhost:8000/api/v1/extract",
        files={"file": ("document.pdf", f, "application/pdf")},
        params={"output_format": "markdown"},
    )

result = response.json()
print(result["content"])

# Direct CLI usage (requires local install)
docflow extract my-document.pdf -f markdown -o output.md

4. Get Raw Text (No JSON Wrapper)

curl -X POST http://localhost:8000/api/v1/extract/raw \
  -F "file=@my-document.pdf" > extracted.md

5. Explore Available Formats

curl http://localhost:8000/api/v1/formats | jq

Response:

{
  "extraction_engines": ["tika", "pymupdf", "auto"],
  "ocr_engines": ["tesseract", "google_vision", "none"],
  "output_formats": ["markdown", "json", "plaintext"],
  "max_file_size_mb": 100
}

Next Steps