Linux
Ubuntu 22.04 and 24.04, Debian 12, Alpine 3.20
ssh dockup@worker-01
Deploy apps, databases, and virtual machines from Git or straight from your AI coding agent.
Works with the tools your agents already use
GitHub deployments
Connect GitHub once. Every approved change becomes a repeatable deployment — built in isolation, verified before it takes traffic, and reversible from the same screen.
Deploy on every push, or hold each release for manual approval.
Dockup watches the branch you nominated. Git stays the source of truth — there is no separate state to keep in sync.
Dependencies cached, variables applied, the image assembled away from whatever is currently serving.
Nothing moves until the new version answers. If it never does, your users never see it — the release stops here.
The new version takes the domain. The previous one is held, so rolling back is a switch rather than a rebuild.
The same twenty seconds, either way
Then it is one service, not six tools
Managed databases
PostgreSQL, MySQL, MongoDB or Redis, provisioned with credentials, TLS and encrypted storage — and reachable only from inside your workspace until you decide otherwise.
Version, region, CPU and memory, and encrypted storage are chosen at creation and changed later without rebuilding.
main-db.internal:5432Your services resolve it by name over TLS, with credentials Dockup generates and injects into the ones that need them.
There is no public hostname to firewall, because there is no public hostname. Open one deliberately when an outside client genuinely needs it.
Provisioning one, either way
Day two, from the same screen
Virtual machines
A whole machine with root on it — Linux or Windows — provisioned in under a minute and already sitting on the same private network as your apps and databases.
Linux
Ubuntu 22.04 and 24.04, Debian 12, Alpine 3.20
ssh dockup@worker-01Windows
Windows Server, full desktop
RDP, on a tunnelled portEither one, on the same canvas
Identical either way
workspace.internalCloudflare domains
Connect a Cloudflare account once, with a token scoped to just two permissions. After that a hostname becomes a live HTTPS endpoint from one line — without opening Cloudflare again.

Those two are the whole grant. Dockup cannot read your other zones, change your account, or touch anything it was not given.
One line, either way
What that produced
Written through the scoped token. Propagation is typically under a minute.
api.acme.com reaches agent-api on port 3000, and keeps following it through redeploys.
Verified and reissued on its own. There is no step where you upload anything.
Remote servers
Your machine stays yours. Register it over SSH, write the steps you already run by hand once, and Dockup executes them in order — keeping the output, the timings and the exit code.
Nothing is installed on the box and nothing migrates. Dockup connects as a user you scope, over a key you provide.
The same task, either way
Simple pricing
The plan buys you the platform. Infrastructure — compute, storage, bandwidth — is metered separately, and no paid plan caps how many workspaces, deployments, databases or machines you create. Free starts you with $10 of it.
$0/ month
Sign up and $10 of infrastructure credit is already there. Enough to run something real before you decide anything.
$5/ month
A full workspace with nothing counted against you.
$20/ month
Everything in Hobby, and the fee comes back as usage.
$20 plan−$20 usage credit=$0if you already spend $20 a month on infrastructure
Choose ProRun Dockup on hardware you already pay for. One command on your machine attaches it to your account, and only your account: nobody else is ever scheduled onto it. You keep the server and its bill; we license the platform and give it a domain.
What “unlimited” means. Dockup does not cap how many workspaces, deployments, databases or virtual machines you create. What you run still consumes compute, storage and bandwidth, and that is metered — Pro simply covers the first $20 of it each month.
Ten reviews from people running production on Dockup. Published in full, in their words.
We started using dockup.ai because we wanted to spend less time dealing with servers and deployment configuration. Connecting our Git repository and getting the application live was surprisingly straightforward. SSL, domains and deployments are handled for us, which saves a lot of time compared to managing everything manually.
The biggest advantage for me is how well Dockup fits into an AI-assisted development workflow. I use Claude Code and Codex regularly, and with the Dockup CLI they can actually interact with our deployments instead of just generating commands for me to run. Being able to deploy, check status and inspect logs from the CLI is a huge improvement.
We previously had several different tools for CI/CD, hosting, databases and monitoring. Dockup has simplified that setup considerably. We can deploy our services, manage environment variables, check deployment history and access logs from one platform. For a small engineering team, that makes a real difference.
The PR preview functionality has become one of our favorite features. Every time we work on a larger change, having an isolated preview environment makes it much easier to test before merging. It has made our development and review process noticeably faster.
I wanted to launch our application without spending days learning Docker, servers and complicated CI/CD configuration. Dockup made the process much easier. We connected our repository, configured the environment and had a working production deployment very quickly. Exactly what I was looking for as a founder.
I initially expected Dockup to be mainly a simple deployment platform, but the CLI is actually quite powerful. Being able to inspect services, deployments, environment variables and runtime logs directly from the command line works very well for our automation. The structured JSON output is especially useful in scripts and agent workflows.
Being able to deploy a database alongside the application is extremely convenient. We use PostgreSQL and no longer have to set up a separate database infrastructure just to get a new project running. It makes spinning up new environments much faster.
We have several applications running through Dockup and the deployment workflow has been very smooth. The zero-downtime approach is particularly important for our production services. We can push updates without having to manually coordinate the deployment process every time.
I build and launch products fairly quickly, so I don’t want to spend my time managing infrastructure. Dockup gives me a simple path from code to a live application. I especially like that I can add a database, configure the domain and deploy without setting up a complicated server environment myself.
What convinced us about Dockup is the combination of a straightforward deployment platform and the ability to work with AI coding agents. Our developers can build with Claude Code or Codex and then use Dockup to handle the deployment and operational side. It feels much closer to how modern development teams actually want to work.
Reviews are published with permission and shown in full. “Verified customer” means Dockup confirmed the reviewer uses the platform — these are not collected by an independent review service.