turbomemLocal-first agent memory for TypeScript

No Python. No servers. Just npm install.

Get started

View on GitHub

Fully embedded

Runs inside your Node or Bun process — no sidecar server, no HTTP hop per memory call.

Semantic search

LLM fact extraction plus vector embeddings with PGlite and pgvector (WASM Postgres).

Adapter-based

Swap OpenAI or local WASM embeddings, PGlite storage, and Mastra or Vercel AI SDK integrations.

Type-safe

Strict TypeScript with Zod-validated inputs. Works in Node 20+, Bun, and serverless.

Terminal CLI

Add, search, and manage memories from your shell with one-shot commands or an interactive REPL.

Quick example

import { TurboMemory } from "turbomem";

const memory = new TurboMemory({
  embeddings: "openai",
  storage: "pglite",
  extraction: { provider: "openai", model: "gpt-4o-mini" },
  openai: { apiKey: process.env.OPENAI_API_KEY },
});

await memory.init();

await memory.add(
  [{ role: "user", content: "I love hiking and I'm training for a half marathon this fall." }],
  { userId: "user_123" },
);

const results = await memory.search("What outdoor activities is the user into?", {
  userId: "user_123",
  limit: 5,
});

console.log(results.map((r) => r.memory.content));