A second brain that follows you

Capture from web, chat, and editor — recall from anywhere.

error_outline

Without memory

Your knowledge is split across a notes app you stopped opening, browser bookmarks you never revisit, and chat conversations you can’t search. The tools that capture well don’t recall well, and none of them talk to your AI.

check_circle

With Rekall

One memory layer captures from everywhere you work and recalls into every tool you think in. Unlike a notes app, it’s queryable by meaning, linked by a real graph, and available to your AI assistants — not just your eyeballs.

How it works

Capture without friction

Browser extension for the web, hooks for your editors, MCP for your chats — capture is ambient, not a chore you abandon.

Memory organizes itself

Categorization, entity extraction, and graph linking happen in the background — your notes app’s tags, minus the tagging.

Ask, don’t browse

Recall by meaning (“that article about pgvector performance”) from any connected surface, including your AI tools mid-task.

See it in action

Capture once, recall by meaning

// Captured in the browser three weeks ago;
// recalled in Claude Code today, mid-task
const hits = await rekall.memories.search({
  query: 'that benchmark about pgvector vs Qdrant',
})
// → the article, your highlight, and the note you
//   added: "their filter benchmarks used HNSW off"