ocrlocal

Dead-simple offline OCR for developers. Point it at an image. Get text. No cloud. No API keys.

ocrlocal wraps Tesseract via pytesseract + Pillow into a polished CLI with batching, JSON output, and a delightful doctor check.

Install

pip install ocrlocal

You also need the Tesseract binary on your system:

Verify everything in one shot:

ocrlocal doctor

60-second demo

# Single image → stdout
ocrlocal receipt.png

# Language pass-through (Tesseract style)
ocrlocal scan.png --lang eng+fra

# Batch + JSON for scripts
ocrlocal photos/*.png --json

# Write .txt files into a folder
ocrlocal invoices/*.jpg --out ./text

Sample JSON:

[
  {
    "file": "receipt.png",
    "text": "TOTAL 12.50",
    "lang": "eng",
    "engine": "tesseract"
  }
]

ocrlocal doctor

A green/red health check for your local OCR stack:

ocrlocal doctor
═══════════════

  ✓  ocrlocal           v0.1.0
  ✓  tesseract binary   /usr/bin/tesseract — tesseract 5.3.0
  ✓  pytesseract        v0.3.13
  ✓  Pillow             v10.4.0

All checks passed. Ready for offline OCR.

If Tesseract is missing, doctor exits with code 2 and points you at apt / brew / choco / the Windows installer.

Features

vs calling pytesseract yourself

Use the library if you are embedding OCR deep in an app. Use ocrlocal when you want a dependable terminal tool.

CLI reference

ocrlocal [OPTIONS] [IMAGES]...
ocrlocal doctor

Exit codes

• 0 — success
• 1 — one or more OCR / file errors (partial batch still reports successes)
• 2 — missing Tesseract or broken config

Troubleshooting

Development

git clone https://github.com/AshSgDe29071999/ocrlocal.git
cd ocrlocal
pip install -e ".[dev]"
pytest

Integration tests that hit a real Tesseract binary are skipped automatically when tesseract is not on PATH.

License

MIT © AshSgDe29071999