Pagenova
Privacy policy
Effective date: August 2026
Pagenova is a browser extension that answers questions with AI models. You can optionally include tabs, documents, selected text, screenshots, and images as context. This policy explains what the extension stores, reads, and sends.
Data stored on your device
- Settings are stored in your browser's
chrome.storage.local. API keys are stored only in chrome.storage.session, which Chrome clears when the browser restarts. They remain on your device except when an API key is sent to the provider you selected.
- Chat history and document or tab index data are stored in IndexedDB (
pagenova-history and pagenova-rag) on your device.
- After you allow Page Context, Pagenova reads open-tab titles and URLs in memory to show the tab picker. It does not store or send them as browsing history.
- The optional in-browser embedding model downloads from Hugging Face on first use and is cached locally. Text used to make embeddings stays on your device.
- Clear stored data in Settings removes local and session data, including history and retrieval data.
Data sent to third parties
Pagenova sends data outside your device only when you start an action that uses an external service.
- Your prompt, together with any page text, selected text, document, or image you add, is sent to the AI provider you choose: a local Ollama server, OpenAI, Anthropic, Google Gemini, or an OpenAI-compatible endpoint you configure. To answer a follow-up about an image, Pagenova may send the most recent image attachment again within that conversation.
- If you turn on web search, Pagenova sends your question to DuckDuckGo or Tavily. For a follow-up such as “search it,” it also sends up to two recent user questions so the search provider can identify the subject. This applies to a local Ollama model too.
Limited use
Pagenova does not sell user data, use it for advertising, or give people access to it. It uses page content only to provide the context feature you selected, consistent with the Chrome Web Store User Data Policy's Limited Use requirements.
What Pagenova does not collect
- Pagenova does not collect analytics, telemetry, crash reports, or tracking data.
- It does not collect account, identity, or browsing-history data beyond the open-tab titles and URLs described above.
- It does not send page text or images unless you select a tab as context, highlight text and send a question, or attach an image.
Permissions
Pagenova requests host access for http://*/* and https://*/* so it can ground an answer in any tab you choose and fetch images you right-click or add from the current page. It reads page text and images on demand. The extension does not run a persistent content script on every page. It injects an extraction function only after you select a tab as context or send a question with a page selection, then discards it. Pagenova uses host access to the selected AI provider or local Ollama server only to make the request you initiated.
When you click Screenshot, Pagenova may ask for optional all-sites screenshot access. If you allow it, the extension can capture the currently active normal webpage after you switch tabs while the side panel stays open. It never captures tabs automatically, and it sends a screenshot to a provider only when you attach it and send a message.
These permissions are used solely for the features described above.
Need help?
Questions about privacy?
Open a support issue and include only the information needed to reproduce the problem.
Get support →