Close the laptop. The agent keeps going.
You know the commute with the lid still open. Cloud execution exists. It never feels like the default. Connect a repo, choose any model, close the lid. The agent keeps writing code.
The laptop can close.
- 1
Connect a repo
Your project is in the cloud. Nothing to leave running on the laptop.
- 2
Start a track
Choose any model. It keeps writing code after you close the lid.
- 3
Share the track
Send a link. Teammates review the plan, not the PR.
Review the plan, not the PR.
Share a track and a teammate can read the plan, the commands, and the files without checking out a branch.
Shared track
ravix.sh/t/persist-cart-session
Maya opened this 12 minutes ago. Same plan. Same files. Same terminal history.
Pro or Max.
Same product on both. Inference included. Plans only change how much you can run. Max is capped at $200. The first 100 we invite lock it in at $100.
Max
First 100
$100/ month$200
Half price, locked in.
The top plan. Keep a sandbox on. Same product as Pro. Capped at $200.
Join the waitlistBefore you join.
Doesn't cloud execution already exist?
It does. People still leave the laptop open, because it never feels like the default. Here, there is no local machine to keep awake.
Which models can I use?
Any model. You don't bring a ChatGPT or Claude subscription. Inference is included — we call the APIs.
What's the difference between Pro and Max?
How much included usage. Every feature is on both: cloud tracks, any model, share a link. Max is the top plan, capped at $200.
What does the first 100 get?
Max at $100 instead of $200, locked. After that, $200.
What do you access?
Only the repos you grant. You choose which ones become projects.
When do I get access?
We're taking a waitlist. We'll email you when it's your turn. No product spam in the meantime.
Join the waitlist.
Tell us where to send the invite. One email when it's your turn.