Your first decision when replacing an open source automation tool is whether you need to host it yourself. If you do, stick to open source or self-hostable options. If you don't, hosted no-code builders are the quickest route, and enterprise integration platforms suit larger IT teams connecting many systems.
Self-hosting is the first fork in the road. Activepieces describes itself as open source at its core, and for many of its users that's the whole point: the flows and the data stay on their own servers. If that matters to you, half the market drops away straight off. If it doesn't, you have a much wider choice. The nine alternatives below, in no particular order, are grouped by how you'd build with them.
Do you need to run it on your own server?
Think about who'll be on the hook when an automation breaks at the weekend. A hosted tool means the vendor keeps the servers running and you fix the flow. A self-hosted one means you do both, in exchange for control over where the data goes and what gets installed.
Open source you can host. Windmill, Automatisch and Kestra each advertise open source on their own pages. The first is code-first and pitched as easy to self-host, the second is a no-code option you can run on your own server or in the vendor's cloud, and the third orchestrates workflows written in a declarative language.
Projects for tinkerers. Node-RED and Huginn suit people who enjoy building. One wires real-time data into flows with low-code blocks, and the other lets you create agents that read the web, watch for events and act for you.
Hosted point-and-click. Make and Zapier are the familiar hosted route: you connect apps, set triggers and build flows visually, with AI help on both. n8n sits between camps, mixing code steps with no-code building blocks for technical teams.
Enterprise integration. Tray.ai is a hosted integration platform aimed at IT teams connecting apps, databases and APIs and putting rules around AI agents.
Visual, code-first or integration platform
How you build matters as much as where it runs. Visual builders are quick for marketing and operations people, but complex logic can turn into a sprawling diagram. Code-first tools let developers use version control, tests and reviews, which pays off once flows become business-critical. Enterprise integration platforms add governance, permissions and audit trails, which a five-person team rarely needs.
Nine automation tools compared
| Tool | How you build | Good for |
|---|---|---|
| Automatisch | Open source no-code automation | No-code business automation, hosted or self-hosted |
| Huginn | Self-built agents | Building helpers that watch the web and act |
| Kestra | Declarative workflow orchestration | Data and ops workflows written in a declarative language |
| Make | Visual no-code automation | Drag-and-drop AI workflows across many apps |
| n8n | Workflow automation with AI | Mixing code steps with no-code building blocks |
| Node-RED | Low-code event-driven flows | Collecting and visualising real-time data in flows |
| Tray.ai | Integration platform (iPaaS) | Connecting apps, databases and APIs with AI governance |
| Windmill | Code-first workflows | Scripts, internal apps and AI agents you can self-host |
| Zapier | Connected app automation | Rule-based automations across a long list of apps |
Every option, from A to Z
Automatisch
Best for: Small businesses wanting no-code automation they can run in the vendor's cloud or on their own server, from a project that describes itself as open source.
Huginn
Best for: Tinkerers who want to build their own agents that read the web, watch for events and take actions on their behalf, as the project describes.
Kestra
Best for: Data and platform engineers who define workflows in a declarative language, on an open source orchestrator the vendor sums up as one orchestrator for every workflow.
Make
Best for: Non-developers building AI and agentic workflows visually across many apps, by prompt, drag and drop or from an AI chat, as the vendor describes.
n8n
Best for: Technical teams who want AI and business process automation together, with what the vendor calls the flexibility of code and the speed of no-code.
Node-RED
Best for: Hobbyists and engineers collecting, transforming and visualising real-time data in flows, through what the project calls low-code programming for event-driven applications.
Tray.ai
Best for: Larger IT and operations teams connecting apps, databases and APIs and governing AI agents, on what the vendor calls an AI-native integration platform.
Windmill
Best for: Developers writing workflows, internal apps and AI agents in code on a platform the vendor says is easy to self-host and can run any internal software.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
Zapier
Best for: Teams who want workflows and agents that follow fixed rules and run the same way every time, which the vendor stresses, across a long list of connected apps.
Rebuild one real flow in two trials
Choose a flow you run today, such as a form submission that creates a CRM contact and posts a message to Slack. Rebuild it in two tools from different groups, ideally one hosted and one you can self-host.
- Note how long it takes to get the first run working end to end.
- Break it on purpose, for example with a missing email field, and see how clearly the error is reported.
- Check whether you can retry a failed run without rebuilding it.
- Look at where run history and data are stored, and for how long.
- Ask whoever will maintain it to read the finished flow and explain it back to you.
My advice for small teams and developers
If no one on your team wants to look after a server, don't self-host just because it's open source. A hosted tool will be calmer. If you do have a developer who's comfortable with deployments, a self-hostable, code-first tool gives you control and version history that point-and-click builders can't match.
For a broader look at the category, see my guide to workflow automation tools and my list of AI workflow automation tools.
Moving your flows across
There's rarely a one-click import between automation tools, so plan to rebuild rather than migrate. List every live flow, mark which ones run daily, and move those first while the old tool keeps running. Switch each one off only after the new version has run cleanly for a week. If flows touch personal data, rules on where that data can be stored vary by country, so check the regulator's guidance before you pick a hosting location.
Do you make workflow automation or integration software? Tool makers can ask to be considered for a featured place on this list.
Want your tool on this list? Featured spots and reviews here are paid partnerships. Here's how to partner with me.
Frequently asked questions
What's the difference between self-hosted and hosted automation?
Self-hosted means you run the software on your own server and look after updates. Hosted means the vendor runs it and you only build the flows.
Is open source automation free to run?
The software may be free to download, but you still need a server and someone's time to keep it updated and secure. Factor that in before you decide.
Do I need to code to use these tools?
Not for the visual builders. Code-first tools expect some programming, though many let you mix code steps with ready-made blocks.
Can automation tools run AI agents?
Many now include AI steps or agents. Start with low-risk tasks, log what the agent does and keep a person in the loop for anything customer-facing.
What is an iPaaS?
It stands for integration platform as a service: a hosted platform for connecting apps, databases and APIs across an organisation, usually with governance and permissions built in.
How do I avoid broken automations going unnoticed?
Turn on failure alerts, check run history weekly and give each important flow a named owner who gets the alert.

