An offline-first memory vault for your ChatGPT history — supercharged with Puter.
Import, search, recall, merge, summarize, and export conversations — all in your browser, with a GitHub-dark inspired UI.
- Offline-first archive explorer: Browse your ChatGPT history directly in the browser.
- Import conversations: Drag-and-drop JSON exports from ChatGPT.
- Fast search: Full-text search across all conversations.
- Semantic recall: AI-assisted retrieval using embeddings + relevance ranking.
- Conversation tools:
- Merge multiple threads into one timeline
- Summarize long chats with Puter’s AI API
- Copy or export to Markdown / JSON
- Modern UI: GitHub-dark inspired interface with keyboard shortcuts.
- Authentication with Puter:
- Continue as guest (local-only)
- Log in / sign up to sync via Puter Web OS
- Sync & Cloud storage (optional): Use
puter.fs.*APIs to persist archives across devices. - AI enhancements:
- Summarization (
puter.ai.chat) - Potential image support (vision and OCR ready)
- Future: voice notes (
puter.ai.txt2speech)
- Summarization (
- Clone the repo and install dependencies:
git clone https://github.com/NeurosynLabs/Memora.git cd Memora npm install - Run locally:
npm run dev
- Open
http://localhost:3000in your browser.
Or try it right away here: Live Demo
- Frontend-only: Everything runs in your browser with IndexedDB for offline storage.
- Worker-powered indexing: A Web Worker handles background search + embedding.
- Puter.js integration:
puter.auth.signIn()→ login / guest sessionputer.fs.write,puter.fs.read,puter.fs.readdir→ optional cloud syncputer.ai.chat()→ summarization and semantic recall
- Zero lock-in: All data is yours; export anytime.
- Vanilla JS + Puter.js SDK
- IndexedDB (offline storage)
- Service Worker (offline PWA support)
- Web Workers (fast indexing & search)
- Tailored dark theme UI
- Multi-model summarization (Claude, Gemini, etc via Puter.ai)
- Richer metadata filters (date ranges, participants)
- Voice note playback with
puter.ai.txt2speech