Incognito AI Hub by Gemini

Simple incognito AI chat powered by Gemini & Local LLM. Free toolkit for translating web pages, text, screenshots, and voice input.

Incognito AI Hub — Private, Memory-Free AI Chat + Voice + Translation Chat with AI, transcribe voice notes, and translate any webpage — with zero conversation history stored. No backend server: data travels directly between your browser and the AI service. ──────────────────────────────── 🆕 What's New in v27 (v26→v27.2.3) ──────────────────────────────── ▶ Local AI Model Support for Text Tasks Image recognition, file parsing, and voice features still require the Google Cloud API. But if your computer is powerful enough, you can now connect Ollama or LM Studio to handle text conversations and translations locally — no data sent to the cloud. ▶ Built-in Read-Aloud in the Reader After translating or processing voice input, the reader can read text aloud. Select a target language and tap "Read Aloud" to listen to any passage. ▶ Smarter Webpage Translation The extension now parses the full page structure — headings, paragraphs, tables, and quotes — before translating, not just screenshots. Academic papers, Wikipedia, and news articles translate with better formatting and accuracy. (We still recommend Chrome's built-in Google Translate for most sites; this feature is best for pages Google Translate doesn't support.) ▶ Sortable Preferred Languages Add up to 3 preferred target languages in Settings (including custom languages). Use the ↑↓ arrows to set priority — your most-used language always appears at the top of translation menus. ▶ Prompt Presets with Writing Guide 7 system prompt presets (A–G) for private chats. Use AI persona settings, target audience, response format, task definition, and output style to build more effective prompt combinations. ▶ Interface in 12 Languages The extension interface now supports 12 languages: Traditional Chinese, English, Japanese, Korean, Spanish, German, Arabic, Portuguese, Russian, Turkish, Ukrainian, and Vietnamese. ──────────────────────────────── 🛠 Core Features ──────────────────────────────── 🔒 Private AI Chat Three independent tabs: Gemini 2.5 Flash (fast and efficient), Gemini 2.5 Pro (most powerful, handles very long documents), and Local Model (Ollama / LM Studio — no API key needed, fully private). Closing the tab clears all conversation history. 7 prompt presets (A–G) with tooltip previews on hover. Gemini supports up to 4 images and PDF uploads. 🌐 Full-Page Webpage Translation One click to capture any webpage and display a complete translation in the reader. Works with news sites, Wikipedia, academic papers (arXiv, PubMed, etc.), and more. 📝 Text & Image Translation Workbench Paste text, or upload / paste a screenshot. AI automatically reads and translates text in the image. 🎙️ Private Voice Notes Record up to 15 minutes (subject to browser memory allocation), or upload an audio file. AI transcribes the speech and identifies different speakers. Results open in the reader — tap any paragraph to translate instantly. 📖 Reader Page Clean, customizable reading interface: • Tap any paragraph for instant translation • Read-aloud button (automatically matches the translation language) • Adjust font size, line spacing, and background color (White / Sepia / Dark) • Save as .txt file ──────────────────────────────── ⚡ Quick Start ──────────────────────────────── 1. Install the extension 2. Open Settings (⚙️) and paste your Google Gemini API key — Or enable local model and enter your Ollama / LM Studio endpoint URL 3. Click the extension icon and choose a tool Get a free Gemini API key at aistudio.google.com ──────────────────────────────── 📋 Technical Notes ──────────────────────────────── • No backend server — all AI requests go directly from your browser to Google Gemini or your local model • Required permissions: activeTab, storage, debugger (used only for webpage screenshot capture) • Built with Chrome Manifest V3 and ES6 module architecture