arkitekt-next

Turn your Python functions into apps you can orchestrate, share, and scale.

---

What is Arkitekt?

Arkitekt is an open platform for building, connecting, and orchestrating computational apps. arkitekt-next is its Python client: a framework that takes your ordinary Python functions and exposes them as remotely callable, orchestratable building blocks — without you having to write servers, APIs, message queues, or UIs.

Annotate a function, run your app, and it becomes available on an Arkitekt server where it can be:

• Called from anywhere — other apps, notebooks, scripts, or the web UI.
• Composed into real-time workflows that wire your functions together.
• Given a GUI automatically, generated from your Python type hints.
• Shared with your team behind central authentication and permissions.
• Packaged and deployed as a Docker container with a single command.

Arkitekt grew out of the needs of data-intensive science (it has first-class extensions for microscopy, imaging, and graph data), but the core is domain-agnostic — any Python workload fits.

📚 The best place to understand the platform and its concepts is the documentation at arkitekt.live.

Installation

pip install "arkitekt-next[all]"

This installs everything, including the arkitekt-next command line interface used to create, develop, containerize, and deploy apps.

Prefer a lean install? Pick only the extras you need:

pip install "arkitekt-next[cli]"            # the CLI + app development tooling
pip install "arkitekt-next[mikro]"          # microscopy / imaging data
pip install "arkitekt-next[kabinet]"        # Docker packaging & deployment
pip install "arkitekt-next[fluss]"          # workflow orchestration
pip install "arkitekt-next[elektro]"       # electrophysiology data
pip install "arkitekt-next[alpaka]"         # want to talk to LLMs? This one's for you.

arkitekt-next requires Python 3.11+ and builds on the asyncio and pydantic stacks.

Quickstart

1. Create an app

mkdir my-app && cd my-app
arkitekt-next app init

This walks you through creating an app and writes a manifest (identifier, version, entrypoint, scopes) into .arkitekt_next/.

2. Register your functions

Any function you decorate with @register becomes a callable building block on the platform. Its arguments and return values are inferred from your type hints — which also drive validation, documentation, and the auto-generated GUI.

from arkitekt_next import register


@register
def greet(name: str, excited: bool = False) -> str:
    """Greet a person by name.

    Args:
        name: Who to greet.
        excited: Add some enthusiasm.
    """
    greeting = f"Hello, {name}"
    return greeting + "!" if excited else greeting

3. Run it

arkitekt-next app run dev

run dev connects your app to a local or remote Arkitekt server with hot reloading — edit your code and the app reloads automatically. When you are ready for production, use arkitekt-next app run prod.

The CLI

arkitekt-next is the command line for all things Arkitekt — one tool across the whole platform lifecycle. It absorbed the standalone arkitekt-server tool, so the same binary that builds your apps also stands up a deployment:

Command group	What it does
app	Build, run and deploy apps from your Python code — scaffold (init), run locally (run dev/run prod), generate typed clients (gen), manage the manifest, inspect, and call functions.
plugin	Containerize your app into flavours and publish it as a deployable plugin.
hub · coord · hubinator · engine	Run the server — the data/compute services, an auth coordinator, the full all-in-one stack, or a standalone deployer.
mesh	Join this machine to the deployment's private WireGuard mesh.
self	Manage your Arkitekt install — upgrade the SDK, print versions, dump diagnostics.

arkitekt-next app init         # scaffold an app
arkitekt-next app run dev      # run it with hot reloading
arkitekt-next hubinator init   # stand up an all-in-one server to run it against

See the full reference in docs/cli.md.

Working with data

Arkitekt automatically serializes and documents standard Python types — str, bool, int, float, Enum, list, and dict. For heavier data (images, arrays, large objects), the platform follows a store-by-reference model: data lives in a central, scalable store and only a lightweight reference travels between apps. Extensions like mikro provide ready-made structures for this, and you can define your own.

See the documentation for details on custom data structures and storage backends.

Documentation & links

• 📚 Documentation: arkitekt.live
• 🧰 CLI reference: docs/cli.md
• 📦 PyPI: pypi.org/project/arkitekt-next
• 🐙 Source: github.com/jhnnsrs/arkitekt_next

License

arkitekt-next is released under the MIT License.