Getting started

1. Sign in

Go to tryharness.ai and sign in with your email. You will get a magic link, no password to remember.

2. Grant screen access

Harness works by seeing your screen, so the first thing it asks for is screen access through your browser. This is a normal browser screen-share prompt, and you can stop sharing at any time. When you are not sharing, Harness simply cannot see your screen.

3. Let the on-device models load

On first run, Harness loads its screen-understanding models directly in your browser. This is a one-time load per device (they are cached afterward). These models run locally, which is what lets your screen stay on your machine.

4. Start chatting

Open the chat window and ask about whatever is in front of you. Some things to try:

• “What is this error telling me?”
• “Summarize what is on this page.”
• “What was I looking at earlier this morning?”
• “Flag me if this deploy goes red.”

Choosing how the AI runs

Harness reaches AI in whichever way you prefer:

• Run a model locally on your own machine. Fully private and free.
• Bring your own key (Venice or Bankr) and spend on your own account.
• Use managed inference, pay as you go. Every account starts with $5 of managed inference free, and you can top up any time.

You can change this in settings under Models and API keys. See Plans for how managed credit and the paid Resident plan fit together.