TableWise is a powerful Chrome Extension that allows you to instantly analyze tabular data from any webpage or your clipboard using Google's Gemini AI, openrouter model or a Local LLM (via Ollama).

• 💰 Financial Research: Instantly summarize quarterly earnings reports or stock history from news sites.
• 🛒 Product Comparison: Compare specs, prices, and online reviews in a clean, query-able format.
• 📊 Sports Analytics: Extract and analyze player statistics from favorite sports websites.
• 🎓 Academic Research: Quickly digitize and question data tables from web-based PDF viewers or reference materials.

• Analyze Page: Automatically detects and extracts tables (HTML or AG Grid) from the active tab.
• Analyze Clipboard: Paste Excel/Sheets data directly to get insights.
• AI Chat: Ask questions, get summaries, and identify trends using the integrated chat interface.
• Persistent Chat History: Previous conversations are saved and restored automatically.
• Agentic RAG: Uses a two-step "Planner & Executor" system to intelligently filter large datasets before answering, ensuring high accuracy.
• Local LLM Support: Connect to Ollama or LocalAI for private, offline analysis.
• OpenRouter Support: Access any model (Claude 3.5, GPT-4, Llama 3) via OpenRouter API.
• Smart Suggestions: Context-aware follow-up questions generated automatically.
• Privacy Focused: Your API Key is stored locally. Data is sent to the AI provider only when you explicitly analyze it.

1. Clone this repository.
2. Open Chrome and navigate to chrome://extensions.
3. Enable Developer Mode (toggle in top right).
4. Click Load Unpacked.
5. Select this folder.

1. Click the extension icon and open Settings (gear icon).
2. Select Provider: Google Gemini.
3. Enter your Google Gemini API Key.
4. Start analyzing!

1. Install Ollama.
2. Important: You must set OLLAMA_ORIGINS to allow the extension to connect. Run Ollama with this command:

   # Mac/Linux
   OLLAMA_ORIGINS="*" ollama serve

3. In TableWise Settings, select Provider: Local LLM.
4. Set Base URL (default is http://localhost:11434/v1).
5. Enter your model name (e.g., llama3).
6. Max Context Rows: Adjust this setting to control how much data is sent to the local model (Default: 50 rows).

1. Get an API Key from OpenRouter.
2. In TableWise Settings, select Provider: OpenRouter.
3. Enter your OpenRouter API Key.
4. Set the Model Name (e.g., google/gemini-2.0-flash-exp:free, anthropic/claude-3-haiku).
5. Click Check Models to see what's available to your key.

• Manifest V3
• Vanilla JS / CSS
• Google Gemini API
• OpenAI-Compatible API Client (Ollama/OpenRouter)
• Agentic RAG Pipeline (Planner/Executor)

MIT