Hylon AI

Hylon is an AI assistant for the work you actually do: thinking something through, writing it down, and shipping the code that comes out of it.

Ask a question and the answer streams back as it is written. Keep related work together in a project, and the documents, code and drafts it produces stay versioned beside the conversation that made them. Pick the thread up again on your desktop or on the web.

WHAT IT DOES
- Chat with a capable model, with streaming replies
- Code assistance: explain, refactor and document, with the output kept as artifacts you can return to
- Projects, so the assistant has the context of a piece of work without you re-explaining it every time
- Skills: packaged instructions the assistant applies to a task, when you turn them on
- Connectors: reach your own tools over the Model Context Protocol
- Plugins: capability bundles that stay inactive until you approve them
- Eighteen languages, with full right-to-left support

YOURS BY DEFAULT
Conversations are stored encrypted on your device. Nothing leaves it until you send it. There is no advertising SDK and no analytics SDK in this app, and nothing you write is sold or used to target you.

Turn on sync and your account carries your work to your other devices. Leave it off and everything stays where you made it. You can delete your account from inside the app at any time, and it is deleted, not deactivated.

BUILT FOR AN OPEN NETWORK
Hylon is built for an open, contributor-powered network. Contributing compute from your device is optional and stays off until you turn it on.

A free account includes a daily allowance. Paid plans raise the limits.