Google Translate API Package

A Python package for interacting with Google Translate services. Supports text (with transliteration), image, document, website, and voice translation.

Three modes available:

• Browser (headless) — default, full features, no visible window
• Browser (headed) — full features with visible Chrome window (enables transliteration)
• Direct HTTP — text-only, no browser required, fastest

Features

• Text Translation: Translate text with auto-detect, transliteration, and suggestions
• Image Translation: Extract and translate text from images
• Document Translation: Translate images and documents
• Website Translation: Translate entire web pages via Google proxy
• Text to Speech: Convert text into audio (direct HTTP, no browser)

Installation

pip install -r requirements.txt

Requires Chrome/Chromium installed for browser mode. Direct mode needs only requests.

Usage

Text Translation

from google_translate import translate_text

# Browser mode (headless) — full features
result = translate_text("Hello", "en", "es")
print(result['translated'])  # "Hola"

# Browser mode (headed) — shows window, enables transliteration
result = translate_text("Привет", "ru", "en", headless=False)
print(result['translated'])  # "Hello"
print(result['target_transliteration'])  # "Privet"

# Direct HTTP mode — no browser needed
result = translate_text("Hello", "auto", "am", mode="direct")
print(result['translated'])  # Amharic translation
print(result['source_lang'])  # "en" (auto-detected)

Image Translation

from google_translate import translate_image

result = translate_image("image.png", "auto", "es")
if result.get('success'):
    with open("translated.png", "wb") as f:
        f.write(result['image_data'])

Document Translation

from google_translate import translate_document

# Images are routed through the image tab
result = translate_document("photo.png", "auto", "es")

# PDF/DOCX files use the docs tab + CDP download
result = translate_document("document.pdf", "auto", "es")

Website Translation

from google_translate import translate_website

result = translate_website("https://example.com", "auto", "es")
print(result['title'])  # Translated page title
print(result['html'][:500])  # Translated HTML

Voice Translation

from google_translate import translate_voice

result = translate_voice("hello", "en", "es")
with open("output.mp3", "wb") as f:
    f.write(result['audio_data'])

API Reference

translate_text(text, source_lang="auto", target_lang="es", mode="browser", headless=True)

Parameters:

• text (str): Text to translate
• source_lang (str): Source language code or "auto"
• target_lang (str): Target language code
• mode (str): "browser" (default) or "direct" (no browser)
• headless (bool): Run Chrome hidden (default True). Only applies in browser mode.

Returns: dict with original, translated, source_lang, target_lang, source_transliteration, target_transliteration, suggestions

translate_image(image_path, source_lang="auto", target_lang="es", mode="browser", headless=True)

Parameters:

• image_path (str): Path to image file
• source_lang, target_lang, mode, headless: Same as above

Returns: dict with image_data (bytes), success, translated_size, original_size

translate_document(file_path, source_lang="auto", target_lang="es", mode="browser", headless=True)

Parameters:

• file_path (str): Path to document or image file

Returns: dict with translation result

translate_website(url, source_lang="auto", target_lang="es", mode="browser", headless=True)

Parameters:

• url (str): URL of the website to translate

Returns: dict with title, url, html, success

translate_voice(text, source_lang="en", target_lang="am")

Direct HTTP, no browser needed.

Returns: dict with audio_data (bytes)

Requirements

• Python 3.6+
• requests
• DrissionPage>=4.0.0 (for browser mode)
• Chrome/Chromium installed (for browser mode)

Notes

• Browser mode uses DrissionPage (CDP protocol), not Selenium
• Direct mode uses Google Translate's internal batchexecute RPC endpoint
• Rate limiting may apply
• For production, consider the official Google Cloud Translation API
• Zombie Chrome processes can accumulate — run taskkill /f /im chrome.exe to clean up