Privane Browser Intelligence

Tracking since Jun 4, 2026.

Changelog

• Jun 9, 2026
  description

  Local on-device WebGPU AI assistant that brings zero-leak intelligence directly into your browser side panel.
  
  Privane is a private-first research companion designed for developers and privacy-focused professionals. It runs a lightweight Gemma 2B model completely offline in your browser, utilizing native WebGPU hardware acceleration. No text, page contents, or API keys are ever transmitted to the cloud.
  
  Key Features:
  • 🧠 100% Sovereign Local Inference: Runs on-device silicon with WebGPU. Zero latency spikes, completely offline-capable, and completely free.
  • 📁 High-Signal Page Summarization: Compress massive developer articles, documentations, and posts into brief high-signal summaries.
  • 🛡️ Secure IndexDB Weight Caching: Automatically caches quantized weights inside standard browser storage for instant sub-second reloads.
  • 💬 Floating Sidepanel Interface: Explain technical concepts, translate complex code blocks, or compile research summaries in real-time.
  
  How it works:
  Unlike traditional cloud-centric AI extensions, Privane spins up an isolated offscreen document context that communicates directly with your tab, parsing text inside your local device boundaries. Enjoy local chats, summaries, and structural edits with absolute context guarantees.

  Your complete local AI workspace. 
  
  Privane is a private-first browser intelligence companion designed for professionals who demand absolute data sovereignty. It runs advanced open AI models completely offline, directly on your hardware. No prompts, page contents, or API keys are ever transmitted to the cloud.
  
  Key Features: 
  • 🔒 Absolute Privacy & Offline Execution: Zero telemetry and total air-gap isolation. Run your assistant anywhere—even without Wi-Fi—with zero latency spikes and $0 monthly cloud costs. 
  
  • 🤖 Browser-Aware Agent Mode: Privane understands your browsing environment. Ask your AI to manage your open tabs, search through bookmarks, or recall your browsing history using natural language. 
  
  • 🧠 Persistent Memory Vault: Build a personalized knowledge graph. Save facts, preferences, and important research notes. Privane remembers what you teach it and retrieves exactly what you need for future conversations. 
  
  • 🔎 Automated Deep Research: Give Privane a topic, and it will analyze the active tab or compile information across multiple sources into a comprehensive, high-signal research document. 
  
  • ✨ Floating In-Line Actions: Highlight any text on any webpage to instantly summarize, rewrite, or explain it. Use the inline "Ask Agent" panel to ask contextual follow-up questions without breaking your focus or switching tabs. 
  
  • ⚡ Global Quick Actions: A persistent floating button lets you instantly capture a new memory, spin up a dedicated project workspace, or log a question from anywhere on the web.
  
  How it works: 
  
  Unlike traditional cloud-centric AI extensions that rely on mandatory subscriptions and external server processing, Privane spins up a secure, isolated environment that communicates directly with your active tab. All data is parsed entirely inside your local device boundaries, giving you a powerful, sovereign AI assistant with absolute privacy guarantees.
  
  21:43

Permissions & access

Permissions

sidePanelstorageoffscreentabsscriptingbookmarksactiveTabhistory

Host access

<all_urls>

Screenshots