# supermemory

> External memory and continuous learning engine for AI agents and assistants

Originally launched as an AI second-brain bookmark manager, Supermemory has evolved into an infrastructure engine providing persistent memory and continual learning for AI agents. It extracts and updates context from documents, chats, and URLs into a vector-graph database, dynamically injecting relevant memories into model prompts via API, MCP, or native integrations.

Website: https://supermemory.ai/

## What it is for

- Provide persistent context and continuous memory across AI agent sessions
- Connect AI assistants like Claude Desktop, Cursor, and ChatGPT via Model Context Protocol (MCP)
- Index and retrieve knowledge from complex documents, raw data, and web URLs for LLMs
- Self-host an air-gapped vector-graph memory layer for enterprise AI workflows

## Features

- Vector-graph database architecture that maps interconnected memories over time
- Learner-1 model that continuously extracts, merges, and updates context
- Real-time token injection into agent context windows
- Model Context Protocol (MCP) support and direct plugins for Cursor, Claude, and ChatGPT
- Advanced document parsing with OCR and diagram extraction for PDFs
- Deployable via cloud API, self-hosted Docker, VPC, or local environments
- Enterprise compliance including SOC 2, HIPAA, and GDPR alignment

## Pricing

Freemium. Free tier available; paid Pro, Max, Scale, and Enterprise plans offered.

## Details

- Type: Web app
- Category: [AI](https://xiry.app/topics/ai.md)
- Open source: yes
- Platforms: Web, API, Self-hosted, macOS, Linux, Windows
- For: AI engineers, developers building autonomous agents, and teams needing persistent context across LLMs.
- Alternative to: Mem0, Zep, LangChain
- Tags: memory, agents, mcp, vector search, knowledge base, developer tools

## Why it was saved

- [@limitlesswithishan, 31 August 2024](https://www.instagram.com/reel/C_Vr94uynGt/): Consolidates scattered bookmarks and notes from Apple Notes, Pinterest, Chrome, Twitter, and Notion into a single searchable hub with an AI writing assistant.

## Compare

- [Alternatives to LangChain](https://xiry.app/alternatives/langchain.md)

## Related

- [Odin AI](https://xiry.app/resources/odin-ai): Enterprise platform to build autonomous AI agents and automated business workflows. (Web app, Paid)
- [Building Effective AI Coding Agents for the Terminal: Scaffolding, Harness, Context Engineering, and Lessons Learned](https://xiry.app/resources/building-effective-ai-coding-agents-for-the-terminal-scaffol): A research paper formalizing architecture and context engineering for CLI coding agents. (Article or course, Free)
- [memary](https://xiry.app/resources/memary): Open-source persistent memory layer for autonomous AI agents using knowledge graphs (GitHub repo, Open source)
- [Parallel Search MCP](https://xiry.app/resources/parallel-search-mcp): Real-time web search and markdown extraction MCP server for AI agents. (Web app, Freemium)
- [ai-engineering-hub](https://xiry.app/resources/ai-engineering-hub): Open-source tutorials and project templates for LLMs, RAG, and AI agents. (GitHub repo, Open source)
- [Hermes 3](https://xiry.app/resources/hermes-3): Open-weights fine-tune of Llama 3.1 built for agentic function calling. (AI model, Open source)

Updated 3 October 2026. Page: https://xiry.app/resources/supermemory

Make your own library: Xiry takes your saved reels, tweets, Reddit bookmarks and Watch Later videos, makes them searchable and pulls the useful stuff out, like this library. It's invite-only for now; join the waitlist at https://xiry.app/#waitlist.