MemoryStack is the AI memory engine that gives your agents persistent recall. Store, search, and retrieve semantic memories with simple REST APIs and SDKs for Node.js & Python.
An AI memory API provides persistent storage and semantic retrieval for AI agents. Unlike traditional databases, it understands context and meaning.
Automatically extract and store facts, preferences, and relationships from conversations. No manual tagging required.
Find relevant memories using natural language. Our vector search understands meaning, not just keywords.
Visualize relationships between entities and concepts. Understand how information connects across your data.
Add persistent memory to your AI agents in minutes. Our SDK handles embeddings, storage, and retrieval—you just make API calls.
import { MemoryStackClient } from '@memorystack/sdk';
const client = new MemoryStackClient({
apiKey: process.env.MEMORYSTACK_API_KEY
});
// Store a conversation as memories
await client.addConversation(
"I prefer dark mode and use TypeScript",
"Great choices! I'll remember that."
);
// Search memories semantically
const memories = await client.search({
query: "What are the user's preferences?",
limit: 5
});
// Returns: [
// { content: "User prefers dark mode", type: "preference" },
// { content: "User uses TypeScript", type: "preference" }
// ]MemoryStack enables seamless memory sharing between agents. Build crews that collaborate with shared context and knowledge.
Transfer context seamlessly when handing off between specialized agents. No information loss, no repeated questions.
Learn about handoffsControl which agents can access which memories. Scope by user, project, organization, or custom dimensions.
Explore scopingNative integrations with popular AI frameworks and LLM providers.
Start free with 1,000 API calls/month. No credit card required.