BasicGPT - Analyze, Summarize & Chat with Any Page

AI page analyzer with local processing. Summarize, extract key points, analyze text. Works offline - 100% private, no data sent.

# BasicGPT - Chrome Web Store Listing (v3.0.0) ## Title (70 characters max) BasicGPT - Privacy-First AI Analysis for Chrome ## Short Description (80 characters max) Local browser AI. Analyze pages offline—no cloud, no tracking, no data sent. ## Full Description BasicGPT is a privacy-first Chrome extension for analyzing and chatting with your browsing content. Pick a workflow, choose your model, and get instant AI insights—all processed locally in your browser. **Two powerful workflows:** - **PageAI:** Analyze web pages with summaries, key points, quick actions, and context chat - **DocumentAI:** Analyze documents with knowledge-focused retrieval and intelligent chat Everything runs locally with WebGPU acceleration. No page content leaves your device. No cloud API dependency. No tracking. ### 🎯 What It Does BasicGPT gives you two focused workflows: **PageAI** - Analyze any webpage in one click - Generate Summaries, Key Points, Topics, and Selected Text insights - Use quick actions like Rephrase, Smart Reply, and Meeting Notes - Chat with full page context for deeper understanding **DocumentAI** - Analyze and chat with your document context - Use knowledge-style chat for retrieval-focused follow-ups ### 🧠 Curated Model Library Choose the model that matches your speed/quality needs: - **Llama 3.2 3B (Fast & Smart)** — ~2.6GB, default model, great speed/quality balance - **Qwen 2.5 7B (Pro Max)** — ~5.1GB, highest quality for complex analysis ### ✨ Key Features **One-Click Analysis** - Right-click menu actions for instant analysis - Full page or selected-text analysis - Custom prompts and saved prompt management **Chat with Context** - Multi-turn chat grounded in the current page or workflow context - Fast follow-up Q&A while you browse **Cross-Tab Intelligence** - View analyzed history across tabs - Cross-tab search for previously analyzed content - Batch tab summarization for multi-page research workflows (Pro) **Long-Page Strategy Controls** - Handles long pages with chunked analysis - Choose full-page (complete) or partial (faster) long-page summarization **Performance & Reliability** - Keep Warm model setting for faster repeated runs - Performance Mode for lighter devices - Progressive model-loading indicators with percentage - Robust token budgeting and dynamic context management per model **Power User Experience** - Keyboard shortcuts (ex: Ctrl+Enter to analyze, Ctrl+L to load page) - Per-tab state isolation so each tab keeps its own analysis - Clean loading/error states and actionable guidance ### 🔒 Privacy & Security ✅ 100% local AI inference in-browser ✅ No cloud AI API dependency for analysis ✅ No telemetry or tracking by default ✅ Data stays on your device ✅ Offline-capable after model download ### ⚡ Performance Snapshot - **First Use:** one-time model download (~2.6GB or ~5.1GB) - **After Cache:** much faster startup and model reloads - **Inference:** optimized with WebGPU + model-aware prompt/token controls ### 🏆 Why BasicGPT Unlike cloud AI assistants, BasicGPT is built for private, local-first browsing intelligence: ✅ Analyze and chat without sending page data to external AI services ✅ Use curated model options for speed vs quality ✅ Switch between PageAI and DocumentAI workflows as needed ✅ Scale from single-page summaries to multi-tab research sessions **Analyze privately. Think clearly.** --- ## v3.0.0 Release Highlights ### New in v3.0.0 1. Workflow-based experience with **PageAI** and **DocumentAI** 2. Curated model lineup with **Llama 3.2 3B** + **Qwen 2.5 7B** 3. Smart quick actions: **Rephrase**, **Smart Reply**, **Meeting Notes** 4. Interactive context chat in side panel 5. Cross-tab analyzed-history view and search workflows 6. Batch summarization for multi-tab research (Pro) 7. Keep Warm + Performance Mode controls 8. Long-page chunk strategy controls (full vs partial) 9. Improved loading UX with model-aware progress details 10. Dynamic token/context management per model capabilities