Cross-model memory
Claude, ChatGPT, Gemini, your phone, your desktop, Slack: one memory.
What it is
Every assistant starts blank, or with a memory only it can read. Nuramem is one memory your assistants share. It isn't chat history or notes. It's a structured record of what matters, one you can see, correct and trust.
Get going
Add the Nuramem connector and sign in once.
Use it when you decide something, a date moves or you meet someone.
Save it in Claude today. Ask in ChatGPT tomorrow. Same answer.
Naming Nura ensures it lands in your cross-model memory, not the assistant's own.
Loads at session start in about 150 ms
Works in Claude, ChatGPT, Gemini, Copilot and Cursor
Never used to train AI
One memory · every surface
In your AI assistant
Loads as each chat starts. Saves when you name Nura. Answers when you ask.
Any MCP client
Connect your assistant →On your devices
Capture a thought with a tap or a keystroke, chat with your memory, get a calm daily nudge.
iPhone — available now · Android — coming soon
Set up your devices →Where your team works
Open a project on any AI and it starts with what your teammates saved. Save a decision in the channel where it was made, or forward the thread. Opt-in only.
Slack · Teams · your Nura address
Bring your team →Your memory in one place
Browse and search everything. Correct what's wrong, confirm what's still true, export it all.
app.nuramem.ai
Open the app →Private and protected
Encrypted, kept separate from everyone else's, never sold. Export anytime · delete means deleted.
Shared projects
Share a project with the people you build with while your personal memory stays private.
How projects work: roles, invitations, privacy →
Free during the launch phase
FAQ
The habit is naming Nura (or Nuramem) plus what you want: "nura, save this", "nura, remember the deadline moved to Friday", "nuramem, what's my latest decision on billing?", "nura, move these to the Acme project." Naming Nura is the part that matters — it makes sure the memory lands in Nuramem, where every model can reach it, instead of in one assistant's private memory.
It saves what matters — not chat logs — and at the start of every session hands the model a compact, current picture of who you are and what you're working on, whichever model you're using.
Name it. Say "nura, save this" when you decide something, a date moves or you meet someone, and Nura saves it to the right place and tells you where. When something is tentative or unclear, it asks first.
You're always in control: you can say "don't save that," browse and delete anything from the web app, and export or erase everything whenever you choose.
Claude (desktop, web and Claude Code), ChatGPT, Gemini, Microsoft Copilot,
Cursor, Windsurf, Zed — any assistant that supports MCP connectors. Beyond
assistants there's the iPhone app, the Mac menu-bar app, Slack, Microsoft Teams, email,
the nura CLI and an API — every one reads and writes the same memory.
Step-by-step setup for each lives on the Connect page.
Yes — Nuramem for iPhone is on the App Store now: share a note from any app or type one, chat with your memory, and get the same calm once-a-day nudge as the Mac app. It's the same memory as every other surface — save on the phone, recall in Claude. Android is coming soon.
Yes. Every account gets a private Nura address (find it in the web app): forward a thread, or cc Nura on the conversation, and the decisions, plans and people in it are saved, with a receipt back in your inbox. Meeting recaps from Zoom or Meet work the same way. One thing to know: everyone on the email is captured too — how email capture works.
Yes — that's shared projects, now available. Create a project in the web app (or just ask your assistant to "create a shared project"), then invite teammates by email and give each a role: Owner (everything, including transfer and delete), Admin (invite, manage members and roles, set the project's rules), Member (read and write) or Viewer (read only). They accept the invitation from the app the next time they sign in with the invited email address — the invitation is tied to the address you send it to, so a teammate who signs in with a different address won't see it.
Once they're in, the decisions, plans and context recorded in that project are recalled by every member's assistant — on whichever model each person uses. Everyone keeps their own private personal memory; only what you put in the project is shared.
No — never. Only what you put in a project is shared, and nothing is copied from your personal memory into one. Nura also keeps personal-life content out of projects even when you name a project in the same breath: a health note, a family matter or money that is your own stays personal, and it tells you so.
An owner or admin removes them (or they leave), and their access ends immediately. What they contributed stays in the project, attributed to them, so the record is whole. A project always has an owner: before an owner can leave, ownership is transferred to someone else.
Yes. Only the owner can delete a project, and for 30 days afterwards it is hidden from every member but not gone — the owner can restore it from the web app. After the 30 days it is purged for good.
In Nuramem's own secure cloud — encrypted in transit and at rest, kept separate from every other user's, and built so saves and recalls are instant. Turning your memory into the picture your assistant loads means our systems read it — so we hold it, and we protect it accordingly. We never sell it and never use it to train AI models.
You stay in control in two concrete ways: export everything anytime from the web app, and delete means deleted — deleting your account erases your records, with backup copies expiring within seven days. Details in the Privacy Policy.
No. For assistants you only ever paste the URL
https://mcp.nuramem.ai/mcp and sign in with your account — Nuramem never
sees your password. Building on the API directly? Sign in to the
web app and get your token under
Account → Data → API access; the API reference is at
https://nuramem.ai/openapi.json.
Free during the launch phase.