Boxes and hosting
Choose where your AI team runs
Use dPanel.ai as the control plane while the Box runs in the environment that fits your ownership, access, and hosting needs.
Set up a Box
- Pick the host. Choose where the Box will run. dPanel.ai connects agents, tasks, channels, Live Context, scheduler, and productions to the host you pick.
- Name it and pick the runtime. The name becomes the Box address under dPanel.ai. Choose the standard dPanel.ai workspace — board, agent controller, and policies built in — or the original Hermes native runtime, and add the Codex coding sandbox when the work needs code execution.
- Connect with your key. An existing instance links with its API URL and API token. Billing stays with you, and dPanel.ai manages orchestration, agents, channels, tasks, and execution once linked.
- Review and pair. The review shows the Box name, address, runtime, controller, and host. A one-time pairing command connects the worker, and the Box goes live when its heartbeat is verified.
Box hosts
dPanel.ai Cloud, your own machine, or your provider account.
On dPanel.ai Cloud the compute comes with your plan and dPanel.ai manages everything. On your own machine, instance, or provider account, the provider bills you and dPanel.ai manages orchestration, agents, channels, tasks, and execution.
dPanel.ai (included in your plan)
You (your own machine or server)
You (your existing instance)
You (your existing Hermes)
You (your host)
You (billed by DigitalOcean)
You (billed by Vultr)
You (billed by Hetzner)
You (billed by Akamai)
You (billed by Vast.ai, marketplace pricing)
You (billed by RunPod)
You (billed by AWS)
One control plane
The controls stay the same wherever the Box runs.
Manage its team, tasks, sessions, schedules, channels, memory, usage, and settings from dPanel.ai.
Run a self-hosted Docker Box on your own computer or server, under your own roof.
Attach an existing remote Box by API URL and token and keep your current hosting.
Create the Box on your DigitalOcean, Vultr, Hetzner, Linode, Vast.ai, or RunPod account, billed by the provider to you.
Attaching a remote runtime, applying a provider key, and raising a Box spending cap each wait for your approval.
FAQ
What is Choose where your AI team runs?
Run a dPanel.ai Box locally on your computer, connect an existing remote runtime, and prepare for supported cloud or partner-host targets.
Who is this dPanel.ai page for?
Use dPanel.ai as the control plane while the Box runs in the environment that fits your ownership, access, and hosting needs.
What does the AI team actually do?
It helps with Run a self-hosted Docker Box locally., Connect an existing remote Box by API or token., Manage its team, tasks, sessions, schedules, channels, memory, usage, and settings from dPanel.ai., while dPanel.ai keeps tasks, approvals, and handoff visible.
Where do approvals happen?
Approval moments include Before attaching a remote runtime, Before applying a provider key, Before increasing a Box spending cap.
What does handoff look like?
Box location, runtime status, attached team, required connections, usage state, and next setup step.