XMem — Intelligent Memory

XMem watches what you type in ChatGPT, Claude, Gemini, Perplexity, and DeepSeek then surfaces relevant memories from your personal memory store before you hit send. Just type “X” on any AI platform of your choice and choose between the three modes Xmem offers to seamlessly store and search your memories, or ask questions about your repository using the Xide feature. Live suggestions As you type in any supported AI chat, XMem performs semantic search against your memory store and renders ghost text inline press Tab to accept, Escape to dismiss. A floating chip displays the count of matching memories. Searches fire only after 8+ characters and only when the cursor is at the end of input. Auto save Your outgoing messages are captured and ingested automatically after the AI finishes responding. /Xingest Queue-based ingestion saves conversations to memory without blocking the UI /Xsearch Auto-inject synthesized memory context when you send a message /Xrepo Browse and query your codebase structure via a file tree panel