Ask your agent to

Use code!

  • From prompt to app, without rebuilding the stack.
  • From laptop to cluster, without rewriting the deploy.
  • From one container to many pods, without re-solving scale.
  • From one agent to a team, without relearning what’s known.
  • From your first user to your billionth, without starting over.
  • From our servers to yours, without lock-in.
  • From one cloud to the next, without rewriting a thing.
  • From someone else’s cloud to your own nodes, without handing over your data.
  • From a fresh host to a working node, without a playbook.
  • From a vulnerable kernel to a patched one, without downtime.
  • From git push to running, without a rented pipeline.
  • From “it’s down?” to “here’s why”, built in.
  • From a domain name to email and DNS, without five SaaS bills.
  • From any agent to the same machines, without a new integration.
  • From a hosted model to your own, without changing the workflow.
  • From side project to paying customers, without building billing.

From burning tokens to compounding them.

The agent writes the code, tests it, builds it, deploys it, and patches it when something breaks. usecode is where all of it runs, on any node, anywhere, configured the way you need it or the way that works best.

The agent builds. You watch it ship.

usecode is open source and freely distributed. Have a look on GitHub.

  1. Joining node-04 to the mesh…
  2. Ingress enabled, routes applied
  3. node-02 drifted, kubelet put back
  4. All 3 nodes match what you asked for

Bring any node

A box under your desk, a VPS, a rack in a shop, or the nodes of your Kubernetes cluster. Each one runs one small daemon, usecoded, and that's all it needs.

We keep an eye on them

usecode.dev supervises your nodes around the clock. When something drifts, a service stops or a node falls behind, it nudges things back the right way, so you don't have to notice first.

A private network between them

Your nodes join a WireGuard mesh, so they talk to each other privately wherever they are: at home, at the edge or in a cloud.

Your agents get the keys

Hand your AI agent an API key and it can reach your machines, run tools and spin up what it needs, in the sentence you already typed.

Open source, freely distributed

usecode is open source and freely distributed under the HGL General License, so you can read the code, run it and share it.

Free for you, no node meter

Your own account costs nothing, whether you add two nodes or two hundred. Teams pay per person, never per node.

Up and running in three steps

  1. 1

    Make an account

    An email and a password. Takes a few seconds, and it's free.

  2. 2

    Hand your agent a key

    One line in Claude Code, Cursor, Codex or whatever you use.

  3. 3

    Add your nodes

    Edge boxes or Kubernetes nodes. usecode.dev takes it from there.

Questions

Is it really free?

For your own account, yes, however many nodes you add. Teams pay $10 per user a month, and for an enterprise setup we work it out together.

What counts as a node?

Any Linux machine you can run usecoded on: an edge box, a VM, a dedicated server, or a node in your Kubernetes cluster.

What happens to my nodes if I cancel?

Your team's nodes keep running exactly as they are. Once the period you paid for ends, your account goes back to Individual.

Can I run it all myself instead?

Sure, usecode is open source. Running it yourself is on your own time; if you'd like a hand setting it up for your team, that's a paid engagement and we're happy to talk.