Screen OCR Translator troch Alex

This add-on uses optical character recognition (OCR) to capture text from any selected area of a web page and translate it. You can switch between the original and translated text, copy it to the clipboard, or listen to it using text-to-speech.

How to use:
Activate the add-on from the browser toolbar, context menu or keyboard shortcut. Select the screen area and click 'Run OCR'. You can also right-click any image and choose 'Translate this image'. The translation is rendered in place on the captured area (default mode), or as a panel in the bottom-right corner.

Keyboard shortcut

Press Ctrl+Shift+F to select a screen region for translation. To change the shortcut, open about:addons, click the gear button, and select Manage Extension Shortcuts.

Features:
- Fully local OCR powered by PaddleOCR
- Automatic detection of the source language
- Instant translation to 70+ languages via Google Translate (no API key required)
- Optional LLM translation through a user-configured OpenAI-compatible endpoint, such as llama.cpp, LM Studio, or a cloud provider
- Text-to-speech through Google TTS. Support for local TTS is planned.
- Support for GPU acceleration (can be enabled in settings).