LLM Sidebar with Context

LLM SIDEBAR: A CONTEXT-AWARE AI COMPANION LLM Sidebar with Context brings the power of AI directly into your browsing workflow. Research, summarize, and analyze multiple sources simultaneously—all from a single side panel without switching tabs or copying and pasting. Use cutting-edge Google Gemini cloud models, tap into hundreds of models via OpenRouter, or keep everything 100% private and offline with a local Ollama model. --- REAL-WORLD USE CASES --- * COOKING & LIFESTYLE: Instantly extract recipe steps and ingredient lists from a cooking video on YouTube. * ACADEMIC & RESEARCH: Synthesize research across multiple tabs or review and draft directly in a Google Doc. * CODING & DEVELOPMENT: Troubleshoot code by pinning Stack Overflow, documentation, and GitHub PRs side-by-side. * SMART SHOPPING & TRAVEL: Pin competing products across multiple stores or compare hotels to generate side-by-side comparisons. --- KEY CAPABILITIES --- * INTELLIGENT CONTEXT PINNING: Don't just chat with one page. Pin up to 6 open tabs to use their collective data as context for your prompts. * DYNAMIC BROWSING: Toggle the "Share Current Tab" icon to automatically include whichever page you are currently viewing in the conversation. * SPECIALIZED EXTRACTION: Native strategies for high-fidelity content extraction from YouTube (video transcripts), Google Docs (document text), and standard web pages. * CHOOSE YOUR PROVIDER: Switch between Google Gemini, OpenRouter, and a local Ollama server anytime right from the sidebar. --- PRIVACY WITHOUT COMPROMISE --- * YOU IN THE DRIVER'S SEAT: Decide exactly which tabs get shared with the AI and when context is cleared. * NO MIDDLEMAN: This extension runs entirely in your browser. Your prompts are sent directly to the Google Gemini API, OpenRouter, or your own local Ollama instance—with zero middleman servers. * FULLY LOCAL OPTION: Run entirely on your machine with Ollama. Point the extension at your local server and your prompts and page content never leave your computer. * LOCAL-FIRST STORAGE: Your API keys, Ollama host settings, and chat history are stored securely in your browser's storage. * OPEN SOURCE: Fully transparent and licensed under Apache 2.0. View or audit the code at: https://github.com/google/llm-sidebar-with-context --- SUPPORTED MODELS --- CLOUD (GOOGLE GEMINI): Fast responses with huge context windows via Google AI Studio: * Gemini 3.5 Flash Lite (Default) * Gemini 3.1 Flash Lite * Gemini 3.8 Flash * Gemini 3.7 Flash * Gemini 3.6 Flash * Gemini 2.5 Pro CLOUD & FREE TIERS (OPENROUTER): Access a huge selection of popular models: * Free Mode (Auto-Updated): Includes openrouter/free router plus the top 5 weekly most popular free models (Llama, DeepSeek, Qwen, and more). * Custom Mode: Search and add any favorite models from the OpenRouter catalog. LOCAL (OLLAMA): Connect to a locally running Ollama server and chat with any model you have pulled (Llama, Mistral, Gemma, Phi, Qwen, etc.). Includes configurable context window size and keep-alive duration. --- IMPORTANT NOTES --- * BYO KEY OR RUN LOCAL: A Google Gemini API key (free at Google AI Studio) or OpenRouter API key (openrouter.ai) is needed for cloud models. To use Ollama instead, run Ollama locally—no API key needed. * DISCLAIMER: This is an open-source community project and is NOT an official Google product. Google specifically disclaims all warranties regarding its quality, merchantability, or fitness for a particular purpose. --- GETTING STARTED --- 1. INSTALL the extension and click its toolbar icon to open the sidebar. 2. CONNECT your provider in Settings: paste your Gemini or OpenRouter API key, or enable Ollama and verify your local host (default: http://127.0.0.1:11434). 3. PIN your context tabs and start prompting!