instibrain
instibrain  /  compare  /  claude memory

instibrain and Claude's memory keep different things.

Claude remembers you, inside Claude. A brain holds your company, for every AI you use. Most teams want both.

Claude's memory is enough for One person who works in Claude.
  • you work alone
  • only in Claude
  • it should remember you
a brain adds What your whole team's AIs read.
  • pages your team shares
  • every AI, not only Claude
  • readers per space
  • sealed from the host
  • files you keep
first, what Claude's memory does better

Claude's memory wins on four counts.

If these are all you need, you need nothing more.

  • It fills itselfClaude saves as you chat. A brain's pages are written.
  • It is part of ClaudeThe free plan included. There is nothing more to sign up for.
  • It knows youYour role, your preferences, the projects you work on.
  • It is yours aloneOn a team plan, an owner cannot read a member's memory.
the differences

Six differences, a picture each.

On the left, Claude's memory. On the right, a brain.

1

What does it keep?

Claude's memory
  • your role
  • your preferences
  • your projects

Saved as topics, from your chats.

What Claude learned about you.
a brain
identity.md  who your AI is
pricing.md   what is true
deploy.md    how the work is done
What your company knows, as pages.
The detail

Claude's memoryClaude builds memory from your chats, as topics: your role, the people in your work, your preferences, and the projects you work on. It saves on its own, or when you say "remember this".

a brainA brain keeps pages your team writes: who your AI is, what is true, and how the work is done. The brain runs no AI and reads no chat: a page exists because a person or their AI wrote it. The brain.

2

Whose is it?

Claude's memory
  • you
  • a teammate
  • your team's owner

Each person has a memory of their own.

One person's.
a brain
team
  • everyone
finance
  • finance
legal · sealed
  • two people
Your team's, each space with its readers.
The detail

Claude's memoryMemory belongs to one person. On a Team or Enterprise plan, an owner turns memory on or off for the organization, and cannot view or edit a member's memory. What a team shares in Claude is a project, on those plans, with people who can view or edit.

a brainEach space has named readers, editors and managers, and it is shared whole, with people, groups and agents. An AI reads only the spaces its person gives it, never more than its person. Spaces.

3

Which AI reads it?

Claude's memory
  • Claude on the web
  • Claude on the desktop
  • Claude on the phone
  • another maker's AI
Claude, in its own apps.
a brain
  • Claude Code
  • Codex
  • Claude on the web
  • Claude on the phone

One address, over MCP. ChatGPT and Cursor speak MCP and have not run on our brain yet.

Every AI you connect, at one address.
The detail

Claude's memoryAnthropic's memory page covers Claude's web, desktop and phone apps. The memory stays with Claude: an AI from another maker does not read it.

a brainAny AI that speaks MCP connects to the brain's address. Run on our brain: Claude Code on Mac, Windows and Linux, Codex, and Claude on the web and the phone. ChatGPT and Cursor speak MCP and have not run on ours yet, so we promise no result for them. Connect your AI.

4

How does it change?

Claude's memory
  1. you chat
  2. Claude saves a topic

You edit or delete a topic in settings.

It fills itself as you chat.Less work: there is nothing to write
a brain
  1. someone edits
  2. the check reads it
  3. a new version

Every version names its writer.

Someone writes it, and a check reads it.
The detail

Claude's memoryClaude saves as you chat, or when you say "remember this". You view your memory in settings, and edit or delete a topic there, or tell Claude in a chat what to change or forget.

a brainA person or their AI edits the page. A check at every save flags a line that is too long, vague, written as history, or repeated from another page. It refuses nothing, so a person still decides. Every version names its writer. Version history.

5

What stays out of reach?

Claude's memory
  • sensitive topics, unless you turn them on
  • an incognito chat

Neither is saved to memory.

You choose what it may remember.
a brain
sealed, on our server
rT9b8QDUyii5i+4ftgFBFd0C1brw==
UqEW2qJ84uwjJFQu4hKyH3oELpNmxG3n==
A sealed space reaches us as ciphertext.
The detail

Claude's memorySensitive topics such as health stay out of memory unless you turn them on, and some details, such as a government ID number, are never saved. An incognito chat is not saved to memory or to your history.

a brainAn open space is not hidden from us: our server reads it, which is how a web AI and a phone read it. A sealed space is encrypted on the computers of the people it is shared with, and we hold ciphertext, so no web AI or phone reads it. We still see its size and timing. The limits, in full.

6

Can you take it with you?

Claude's memory
  1. export from Claude
  2. import elsewhere

Anthropic calls this feature experimental.

An export, marked experimental.
a brain
~/instibrain/spaces/team/
pricing/floors.md
runbooks/deploy.md
The folder is already on your computers.
The detail

Claude's memoryMemory exports from Claude and imports into it, for a backup or a move. Anthropic calls the feature experimental. Your memory is also part of a data export.

a brainThe sync client keeps every page as a Markdown file in a folder on your computers. There is nothing to export: the files are there, and one setting keeps the folder in your own Git too. Linked computers.

side by side

The whole comparison in one table.

Each row is a difference above; its detail is there.

The questionClaude's memoryinstibrain
What it keepsWhat Claude learned from your chatsPages your team writes
Whose it isOne person'sThe team's, by space
Which AI reads itClaude's own appsEvery AI you connect over MCP
How it changesIt saves as you chatWritten, then checked
Sharing with a teamA project, on team plansA space, with three roles
Out of reachSensitive topics, incognito chatsA sealed space
Taking it with youAn export, experimentalFiles on your computers
PricePart of ClaudeFree for one person; a team by request

The facts about Claude are Anthropic's own, read on 10 November 2026: chat search and memory and projects. Check them there before you decide.

which one

Choose by who has to know it.

If only you and Claude need it, Claude already keeps it.

choose instibrain ifYour team's AIs should all know the company.
  • Your team uses more than one AI: Claude for some work, Codex for other work
  • A fact has to be the same for everyone who asks
  • Some pages are for a few people, or for no server to read
  • You want to read and correct what your AI reads
Claude's memory is enough ifYou work in Claude, and it should remember you.
  • You work alone, and only in Claude
  • You want today's chat to follow last week's
  • What it remembers is about you, not about the company
  • Your team works only in Claude, and a shared project holds what it needs
together

Use both: Claude remembers you, the brain holds the company.

Leave Claude's memory on. Connect Claude to the brain, and it also reads your company's pages, as you.

  • Claude's memoryHow you like an answer written, and what you were doing last week
  • The brainThe pricing floor, a client's history, and the deploy runbook, one current page each
  • Your teammate's AIReads the same pages, in the spaces shared with them, whichever AI it is
questions

What people ask who use Claude's memory already.

Does instibrain work with Claude?
Yes. Claude Code on Mac, Windows and Linux, and Claude on the web and the phone, have run on our brain. Connecting Claude Code is one command.
Should I turn Claude's memory off?
No. Claude's memory keeps what is about you, and the brain keeps what is about your company. Neither replaces the other.
Does instibrain learn from my chats?
No. The brain runs no AI and reads no chat. A page exists because a person wrote it, or because their AI wrote it and named itself as the writer. That is more work than a memory that fills itself, and it is why a page can be trusted by the next person who reads it.
Is a shared Claude project the same thing?
A project keeps files and instructions for chats in Claude, and a Team or Enterprise plan shares it with people who can view or edit. A brain keeps pages for every AI you use, each space with its own readers, and the pages are files on your computers.
What does instibrain not do?
It runs no AI and imports no documents. We host the brain: you cannot run it on your own servers, and its source is not published. A personal brain is free, up to 200 pages and 20 MB; a team's price is by request.