01
A memory that stays.
Local RAG connects conversations, notes and files. Projects have a memory of their own, separate from your personal context.
DiscoverMEMO / A LIVING MEMORY01 — 04
OPEN SOURCE. ON YOUR MACHINE.
Your personalAI agent.
Memo remembers your context and gets things done. Your models, tools and skills, working together on your own computer.
or install from the command lineTerminal
curl -fsSL https://data.memocpp.com/get-memo.sh | bashScroll to watch a thought take root
01
A SECOND BRAIN. A PAIR OF HANDS.
More than a conversation. A growing connection.
Memo is an AI harness: the home for your models, memory, tools and skills. Ask it to research, write code or work with your files. It keeps the context and carries out the steps, with permissions you control.
Explore the documentation02
Your models. Your choice.
Run a model locally with llama.cpp, connect an API, or use a supported subscription. Choose how you work, one model at a time.
Discover03
From thinking to doing.
Files, code and everyday tasks, with tools, skills and sub-agents. Plan your work or hand Memo a checklist.
Discover04
A day that feels lighter.
Mail and Digital Twin bring replies, commitments and contacts together. Shape a daily briefing and choose where you receive it.
Discover05
Close, wherever you are.
Keep the conversation going with WhatsApp, Telegram, voice and the mobile companion connected to your own Memo.
Discover06
Control, from the roots up.
Manage permissions, choose external services and use encrypted backups. Official builds include usage sharing you can switch off.
DiscoverA SPACE THAT IS YOURS
A tab closes. A connection stays.
Pick up where you left off. With your models, your memory, your own way of working.
faq
Questions you probably have.
Which computers does Memo run on?
Windows 10 and 11, Linux (Debian, Ubuntu, Fedora, Arch) and macOS (Apple Silicon and Intel), plus an Android companion app. Memory requirements depend on your model; a GPU is optional.
Is Memo really free?
Yes. Memo is open source under the GNU AGPL v3. There is no subscription and no account.
Where does my data go?
Chats and memory are stored locally. Connected providers process the context you send. Official builds share a content-free daily usage summary by default; switch it off in Settings → Privacy. A separate one-time install count is also sent.
Does Memo work offline?
Yes. With a local model running through the bundled llama.cpp engine, chat can work offline. Connect an external provider only when you want more power.
Why does my antivirus warn about the installer?
New apps with few downloads can trigger SmartScreen or antivirus warnings. Memo is open source, so you can read the code or build it yourself.
Can I use Memo with Ollama?
Yes. Connect Ollama as a provider and use Memo’s memory, chat and agent interface with your models. You can also run local models through the bundled llama.cpp engine.
SOMETHING GOOD STARTS SMALL.
Give your thoughts a place to grow.
Free and open source. At home on your machine.
or install from the command lineTerminal
curl -fsSL https://data.memocpp.com/get-memo.sh | bash