I built LumiFlow because of three frustrations with AI chats: 1. Long conversations confuse AI. I'd mention using a tool to build my product, and later it would insist that tool WAS my product. 2. AI remembers things I don't want, but I can't find or delete them. 3. Existing tools charge for premium features and export data in terrible formats. So I built my own solution. LumiFlow compresses conversations at 10:1 ratio, keeping only what matters: goals, decisions, examples, next steps. You control what context the AI sees. It works across conversations (start fresh without losing context), across platforms (ChatGPT to Claude to Gemini), and across sessions (close browser, come back tomorrow). What I'm proud of: - Completely free, no premium tier - 100% local - your data never leaves your browser - Open source (AGPLv3) - audit every line on GitHub So, what context problems frustrate you most when using AI? 🤔

LumiFlow

Save AI context, continue anywhere - new chats or platforms

Save AI context, continue anywhere - new chats or platforms