IN DEVELOPMENT · MAC FIRST
A personal AI
that grows with you.
We’re building Sevra: one continuous conversation for your thoughts, your journal, and the things you want to make. At home on your computer.
macOS first. Windows and Linux to follow.
Chapters
A slower morning
The garden project
Things worth keeping
Your journal
TODAY · A SLOWER MORNING
Let’s keep it simple. A place for the book, a thought you want to remember, and nothing to keep up with.
A little reading journal
Your books. Your notes. At your own pace.
Preview before adding to your space
Less starting over.
More picking up
where you left off.
One conversation, many chapters.
Talk without deciding where a new chat belongs. Sevra is being designed to organize the conversation into chapters you can find, revisit, and edit.
A journal that stays connected.
Keep a thought, work through an idea, or look back on what changed. Your knowledge should grow with the conversation, in plain files you own.
Small tools, made for your way of doing things.
The plan: skills and mini-apps you can shape through conversation. Their data lives in your db.md knowledge base, even when the app changes. You decide what they can access.
Personal means yours.
We’re building for local operation: your AI and personal files on your computer, usable offline after the model download. No Cloud account required for the local app. Sync and backup will be optional.
Built on Slotstream.
The inference engine already runs large mixture-of-experts models by streaming experts from SSD. Sevra is the personal app we’re building on top. The engine works today; the app is still in development.
Explore the engine and measurementsA few things to know.
Can I download Sevra yet?
Not yet. Join the waitlist for development updates and early-access announcements. We haven’t announced a release date or price.
Will it run on my computer?
We’re starting with Apple Silicon Macs, then Windows and Linux with AMD and NVIDIA support. Exact memory, storage, and supported devices will be published after testing. Not every computer will be able to run the local model.
Do I need to choose or set up a model?
Sevra is being built around a carefully chosen model, with setup handled by the app. The goal is a personal app you use, not a catalog of models you manage.
What about the existing Sevra Cloud?
It stays available. Its hosted brains, account access, and existing plans are separate from the new local app.
There’s more to come.
Follow along as Sevra takes shape.
Join the waitlistLooking for the hosted service? Visit Sevra Cloud.