Immersive Translate

Renders original and translated text in a vertical split-layout format by parsing the webpage DOM to identify paragraph-level content blocks, then injecting translated text alongside original content without disrupting page layout or interactivity. Uses intelligent content-area detection to distinguish main text from navigation, ads, and metadata, enabling readers to compare source and target languages in real-time as they scroll.

Detects mouse hover events over paragraph elements and displays translated text in a floating tooltip or inline expansion below the hovered paragraph, preserving surrounding context and allowing readers to selectively translate specific passages without translating the entire page. Implements event delegation on the DOM to minimize performance overhead and only triggers translation API calls for explicitly-hovered content.

Implements privacy-preserving translation architecture by encrypting translation requests using SSL/TLS and proprietary APPI protocol, ensuring translated content is not retained by extension or translation service providers, and not used for AI model training. Provides transparency reports and GDPR compliance documentation, with optional local-only translation mode (if offline translation engine available) for maximum privacy.

Integrates with web-based meeting platforms (Zoom, Google Meet, Microsoft Teams, etc.) by capturing audio streams or subtitle tracks, performing speech-to-text transcription (if needed), translating transcribed text in real-time, and displaying translated captions in meeting interface. Supports speaker identification to attribute translations to correct participant, and enables per-participant language preferences (e.g., participant A sees English, participant B sees Spanish).

Maintains local cache of translated content with source text, target text, translation service used, and timestamp, enabling users to search translation history, review previous translations, and avoid re-translating identical content. Supports cloud synchronization (Pro tier) to sync translation history across devices, with optional privacy controls to exclude sensitive content from cloud storage.

Analyzes webpage DOM structure using heuristics (text density, semantic HTML tags, visual layout) to identify main content areas and exclude navigation, advertisements, sidebars, and metadata from translation. Implements machine learning-based content detection (if available) to improve accuracy on complex layouts, with user override capability to manually mark content areas for translation or exclusion.

Intercepts text input into web form fields (textareas, input[type=text]) and provides real-time translation suggestions or auto-translation as the user types, enabling multilingual form submission and chat interfaces. Detects input events, batches keystrokes to avoid excessive API calls, and displays translated text in a secondary input field or dropdown, allowing users to select which language version to submit.

Processes uploaded or locally-stored PDF files by extracting text content while preserving spatial layout information, translates extracted text using selected translation service, and re-injects translated text into the PDF at original positions, maintaining font sizes, margins, and page structure. Supports export as bilingual PDF (original + translation side-by-side) or single-language translated PDF, with optional OCR for image-based PDFs.

Intercepts subtitle tracks from video players (YouTube, Netflix, etc.) by accessing the video element's TextTrack API or subtitle stream, translates subtitle text while preserving timing metadata (start/end timestamps), and re-renders translated subtitles in the player's native subtitle display. Supports multiple subtitle tracks and allows users to toggle between original and translated subtitles without reloading the video.

Detects images on webpages containing text (via heuristics or user selection), applies optical character recognition (OCR) to extract text from image pixels, translates extracted text using selected translation service, and displays translated text as an overlay or replacement on the original image. Supports both automatic image detection and manual image selection via context menu or click-to-translate.

Provides unified API abstraction over 20+ translation services (DeepL, Google Translate, OpenAI, Claude, Anthropic, Microsoft Translate, Tencent, DeepSeek, Gemini, etc.) by normalizing request/response formats, handling authentication (API keys, OAuth), and implementing automatic fallback to alternative providers if primary service fails or rate-limits. Users select preferred service via dropdown, with optional automatic provider rotation for load balancing or cost optimization.

Maintains curated configuration profiles for popular websites (Google, Twitter, Reddit, YouTube, Bloomberg, Wall Street Journal, etc.) that optimize content detection, paragraph segmentation, and translation quality for site-specific HTML structures and content patterns. Automatically applies site-specific rules when user visits recognized domain, including custom CSS selectors for content areas, exclusion patterns for navigation/ads, and terminology libraries for domain-specific vocabulary.

Processes ePub and other eBook formats by extracting text content while preserving chapter structure, formatting, and metadata, translates extracted text using selected translation service, and exports as bilingual eBook with original and translated text interleaved or in parallel columns. Supports multiple eBook formats (ePub, MOBI, PDF) and maintains compatibility with eBook readers (Kindle, Apple Books, Kobo, etc.).

Maintains user-defined and AI-generated terminology libraries that map domain-specific terms to preferred translations, enabling context-aware translation that respects user terminology preferences and avoids mistranslations of specialized vocabulary. Integrates with translation pipeline to inject terminology constraints into translation requests (where supported by service API), and provides UI for users to curate, import, and export glossaries.