nanoda

nanoda is a side-panel AI assistant that helps you understand and act on the web page in front of you. By default it runs on Chrome's built-in Gemini Nano, so it works immediately with no account to create. When you want a stronger model, connect an OpenAI-compatible endpoint: a cloud service, or a local Ollama / LM Studio / vLLM. ■ On the page - Understand this page: produces a summary, topic tags and suggested questions, and you can keep asking about the page - Fast translate: replaces the text on the page using the browser's built-in translation engine - Summarize / translate / explain / rewrite selected text, also available from the right-click menu - Verify online: for questions that need looking up, sources are cited and the lookup runs automatically ■ Documents - Import a PDF, and relevant passages are cited automatically when you ask - Click a citation to open the original PDF at the exact passage - Check specific documents to scope the conversation to just those ■ Chat - Screenshot, upload and paste images - Chat history is kept locally, searchable and resumable - Set a persona and global instructions to shape the tone of answers - A usage dashboard shows token usage and cost estimates ■ Privacy - Conversations, documents and settings stay in your own browser - With the browser built-in AI model, your information does not leave your device ■ Requirements - Chrome 138 or later - Gemini Nano requires Chrome's hardware support; use a custom node where it is unavailable