You really can get an app up and running in just minutes on

But if you want to know a bit more about what’s what on the platform, then read on. You can skip to the glossary for the tl;dr.

Two big concepts to know about how works, and one small one: Fly Machines, Fly Launch, and Fly Apps.

Fly Machines: The basic unit of Doing Stuff on

Fly Machines are fast-launching virtual machines (VMs) that are individually runnable and controllable. We build Machines out of the images you give us and we run them where you tell us to.

Machines are the compute on, our lowest level of orchestration. When you just have to run a VM of a specific size in a specific place, Machines are what you want. If you’re thinking about individual VMs, rather than applications comprising groups of VMs, then you’re thinking about Machines.

Learn more about Fly Machines.

Fly Launch: Manage applications built on Fly Machines

Fly Launch is how you manage the whole lifecycle of an application on top of Machines, from starting to scaling to changing and redeploying. If you’re not thinking of an individual VM, but rather of an application, like “”, then you’re thinking of Fly Launch.

Fly Launch is our built-from-scratch-for-Fly-Machines orchestrator:

  • Create your app with the fly launch command. Fly Launch detects your framework and gives you useful defaults.
  • Configure your app’s deployment and services with the fly.toml configuration file.
  • Deploy your app’s Machines as a group with the fly deploy command.
  • Scale your app’s Machines horizontally or vertically with the fly scale command.

Again, Fly Launch is built on Machines: you can use Fly Launch to manage the scale of your application with a single command, or you can interact directly with Machines for fine-grained control.

Learn more about Fly Launch.

Fly Apps: We’re not here to tell you what to build

Fly Apps are how we group Machines on the platform. A Fly App can be a web app, or a database, or a bunch of task Machines, or whatever you want to deploy. When we talk about “your app” in our how-to docs, we’re talking about a Fly App.

You can manage a Fly App’s Machines as a group with Fly Launch features, or you can run and manage individual Machines with the Machines API or with fly machine flyctl commands. You can even do all of these things in one Fly App.

Learn more about Fly Apps. glossary

The platform: All the primitives, products, and features that make up the public cloud.

Fly Apps: The way Machines are grouped for admin and management on the platform.

Fly GPUs: Machines, but with GPUs. They boot up with GPU drivers installed and you can run nvidia-smi right away.

Fly Launch: Our orchestrator that includes some good stuff for app hosting, like the fly launch command to get started, fly.toml for configuration, the fly deploy command to deploy all your app’s Machines into new versions/releases, and the fly scale command to scale Machines.

Fly Machines: Firecracker microVMs that launch quickly in any region supported by A VM, or virtual machine, functions like a physical computer, but is software-based. Multiple VMs can run, completely isolated, on one physical host. If you’ve deployed an app on, then it’s running on Fly Machines. There’s a fast REST API to manage Machines, but you can also use flyctl—the Fly CLI—to manage everything from the command line. And then there’s Fly Launch, which combines flyctl commands with a shared config to manage your app’s Machines as a group.

Fly Volumes: Local persistent storage for Fly Machines. Every Fly Volume can be attached to one Machine at a time and belongs to one Fly App.

Organizations: Administrative entities on that let you to manage billing separately, control access by adding and removing members, and share app development environments.