Incognito PrivoAI

Voice-to-text, AI chat, web translation, AI reading. Gemini API key or local AI — no backend, data stays in browser.

──────────────────────────────── 🆕 What's New in v28.2.1 ──────────────────────────────── ▶ Model Refresh: Gemini 3.1 Flash Lite Now Official The default model graduates from preview to stable release for improved reliability. The Gemini 2.5 Flash tab is replaced by Gemini 3.5 Flash, applied across all chat tabs, translation, and settings. Gemini 3.1 Pro (preview) remains available for switching. ▶ Voice Engine Upgrade: New TTS Model & Voice AI reading and voice summary now use the updated Gemini TTS model. Default voice changed from Aoede to Kore — warm, steady, and clear. ▶ Popup Redesign: Direct Access to Settings & Privacy Settings and Privacy Policy entries added to the popup launcher so first-time users can find API keys, AI mode, language preferences, and privacy details at a glance. ▶ Fixes & Cleanup Model name strings updated across all 14 locales; minor bug fixes and code cleanup. ── Key Features ── 🎙️ Voice Notes Record audio or upload a file — AI transcribes speech and labels different speakers automatically. Tap any paragraph to translate it instantly. Great for meeting notes, interview transcripts, and multilingual memos. 💬 Incognito Chat Open a private Gemini AI conversation that leaves no trace after closing. Four tabs for different models: Gemini 3.1 Flash Lite — fastest, great for everyday questions and translation Gemini 3 Flash — balanced speed and quality; supports image and document uploads Gemini 3.1 Pro (preview) — deepest reasoning for long documents and complex analysis Local Model — connect Ollama/LM Studio for fully offline text chat Not sure which to pick? Flash Lite covers most everyday needs. Supports 7 custom prompt presets. 🌐 Translation Workbench Paste text, drop a screenshot, or translate the current webpage directly. Supports OCR for image text, lets you choose a target language, and includes an AI summary option. 🔊 AI Reading & Voice Summary [Beta] Convert YouTube captions, audio files, webpage content, or pasted text into natural speech audio or a spoken summary. Powered by the updated Gemini TTS (Kore voice, warm and steady). Supports custom reading instructions; outputs WAV. 📖 Web Reader & Translation Renders complex webpages as a clean reading view, complementing Chrome's built-in translate for pages it can't handle or when an AI summary is needed. Built-in voice playback: select a language and tap "Read Aloud" to listen paragraph by paragraph. ── Privacy by Design ── No backend server: all AI calls go directly from your browser No data collection: your inputs and outputs are invisible to the developer Self-managed key: Gemini API key stored only in Chrome local storage Local AI support: connect Ollama/LM Studio for text chat and translation with zero network exposure ── 14 Interface Languages ── Traditional Chinese, English, Japanese, Korean, German, Spanish, Arabic (RTL), Russian, Portuguese, Portuguese (Brazil), Turkish, Ukrainian, Vietnamese, Czech. ── Getting Started ── Install the extension Visit Google AI Studio (aistudio.google.com) to get a free Gemini API key Click the extension icon → Settings → paste your key Start using all features!