# hippocampOS

> hippocampOS is a private AI memory layer that turns authorized email, meetings, notes, and files into connected context for Claude, ChatGPT, Claude Code, and compatible MCP clients.

People organize memory into named brains so context for different roles, projects, or teams stays intentionally scoped. Use the links below for the current product description, connection instructions, technical details, security boundaries, and policies.

## Product

- [Home](https://hippocampos.io/): Product overview, use cases, supported sources, and how named brains work.
- [Connect hippocampOS](https://hippocampos.io/connect): Setup instructions for the read-only hippocampOS MCP connector in Claude, ChatGPT, and Claude Code.
- [hippocampOS for business](https://hippocampos.io/for-business): Shared-memory workflows for teams.

## Research and trust

- [Technical whitepaper](https://hippocampos.io/whitepaper): How hippocampOS versions source evidence, constructs memory, and retrieves grounded context.
- [Retrieval benchmark](https://hippocampos.io/benchmark): Reproducible retrieval results and benchmark methodology.
- [Security](https://hippocampos.io/security): Connector authorization, scoped recall, encryption, deletion controls, and documented security boundaries.
- [Privacy statement](https://hippocampos.io/privacy): How account, connector, memory, AI, MCP, billing, analytics, and operational data is handled.

## Policies and support

- [Terms and conditions](https://hippocampos.io/terms): Terms covering accounts, connectors, imports, memory features, subscriptions, and service use.
- [Refund policy](https://hippocampos.io/refunds): The seven-day refund policy and request process.
- [Accessibility report](https://hippocampos.io/accessibility): Accessibility targets, safeguards, test evidence, and open review boundaries.
- [Contact](https://hippocampos.io/contact): Privacy requests, account questions, support, and product feedback.

## Optional

- [Articles](https://hippocampos.io/articles): Practical writing about private AI memory, grounded context, and source evidence.
- [Claude Memory comparison](https://hippocampos.io/compare/claude-memory): How Claude Memory differs from and works alongside hippocampOS.
- [ChatGPT Memory comparison](https://hippocampos.io/compare/chatgpt-memory): How ChatGPT Memory differs from a source-linked hippocampOS memory graph.
- [hippocampOS vs Claude Memory](https://hippocampos.io/articles/hippocampos-vs-claude-memory-2026): A source-backed comparison of Claude's native memory, connectors, and portable memory across AI clients.
- [Mem.ai alternatives for persistent AI memory](https://hippocampos.io/articles/mem-ai-alternatives-persistent-memory-claude-chatgpt-2026): A current comparison of Mem, hippocampOS, Mem0, Notion MCP, and native AI memory.
- [hippocampOS vs Mem](https://hippocampos.io/articles/hippocampos-vs-mem-best-ai-memory-app-2026): A source-backed comparison of their data models, integrations, permissions, and AI-client access.
- [Why useful AI memory starts with evidence](https://hippocampos.io/articles/why-useful-ai-memory-starts-with-evidence): Why trustworthy memory should preserve a path back to its source.
- [Persistent memory tools for ChatGPT and Claude](https://hippocampos.io/articles/best-tools-persistent-memory-2025): A comparison of approaches to persistent AI memory.
- [One memory layer, many AI tools](https://hippocampos.io/articles/one-memory-layer-many-ai-tools): Why working context should remain portable across AI tools.
- [How to organize memory for AI](https://hippocampos.io/articles/how-to-organize-memory-for-ai): A practical system for separating and grouping AI memory.