Skip to content

Activepieces

An open-source automation builder you can self-host, connecting apps into flows the same way Zapier does, with MCP support for wiring in AI agents.

Automation & workflows Agents

Checked

Open activepieces.com

What you'd use it for

Automating tasks across apps the way Zapier does, with the option to run the whole thing on your own servers instead of a vendor's.

  • Trigger a flow off a schedule, a webhook, or an event in one of 760+ connected apps
  • Let an AI agent (Claude, ChatGPT, or your own) call an Activepieces flow as a tool over MCP, rather than only firing on a webhook or a timer
  • Run an AI agent as a step inside a flow to handle a task that needs judgement, not just a fixed action
  • Store and manage structured data in built-in Tables without standing up a separate database
  • Self-host the entire platform on your own infrastructure via Docker Compose or a Kubernetes Helm chart, keeping automation data off a vendor's servers

The verdict

The self-hosting option is real, not a checkbox — you get the same MCP and Agents features as the cloud plan on your own infrastructure for the cost of a server. Credits burn faster than the free tier's daily allowance suggests once a flow calls AI steps, so treat Free as a proof of concept and budget for Plus once anything runs in production. It suits a small business or a technical operator who wants Zapier-style automation without customer data leaving their own servers; it does not yet match Make or n8n for sheer app-library depth and polish.

What it costs

Tier Price For Includes
Free $0 A first automation, or a self-hosted deployment with light usage. 100 credits a day, unlimited flows, Flows/Agents/Tables/MCP access, API access.
Plus $20/month (annual billing saves roughly 2 months a year) A small business running a handful of production automations. 10,000 credits a month, up to 5 users, bring-your-own AI keys, team analytics, API/SDK access.
Team $200/month (annual billing saves roughly 2 months a year) A team standardising automation across several departments. 50,000 credits a month, 25 users, unlimited projects, SSO with standard roles, global connections, email support.
Ultimate Custom, sized to a yearly credit pool Larger organisations that need dedicated infrastructure and audit controls. Unlimited users, dedicated workers, execution priority, SCIM and custom roles, Git sync, audit logs, secrets management.
Embed From $36,000/year A software vendor white-labelling automation inside its own product. White-label builder, SDK access, custom integrations.

Read from Activepieces's own pricing page on 25 Sep 2026. Vendors change prices without notice.

What you need to start

  • Cloud: a free account, no card required for the free tier
  • Self-hosted: your own server with Docker Compose v2 (Postgres and Redis are installed alongside it by the one-line setup script), or an existing Kubernetes cluster for the Helm chart route

A cloud account and a first flow take minutes; a self-hosted instance is one shell command against Docker Compose, but budget longer if you don't already run a server to put it on.

Your first run Sign up on the free cloud plan, click New flow, pick a trigger and a piece from the library, test it, and publish — a working flow is a handful of clicks away, not a project.

Where it breaks

Cloud-hosted or self-hosted (Docker/Kubernetes) — the self-host option, unusual among mainstream automation tools, is the main reason people choose it over Zapier or Make when data residency matters.

First entry in this reference. There is no previous check to compare it against.

When not to use it

Skip it if you want the biggest library of prebuilt integrations and don't care where the data sits — Make and Zapier are more mature there. It also isn't the right pick for a non-technical team that has no one willing to run `docker compose` and keep a server patched, since self-hosting is the whole reason to choose it over a cloud-only competitor.

Activepieces is what you reach for when the appeal of Zapier is real but the idea of your customer data routing through a third party’s servers is not. It’s open source, and the self-hosted deployment is a genuine option rather than a checkbox nobody uses — a business with a compliance requirement can run the whole thing on its own infrastructure for the price of a server.

The building blocks are familiar if you’ve used any flow-based automation tool: triggers, actions, branching logic, and a growing library of prebuilt “pieces” for common apps. What’s newer is MCP support, which lets an AI agent (Claude, ChatGPT, or your own) call into an Activepieces flow as a tool, rather than the automation only ever being triggered by a webhook or a schedule. That’s the wedge separating it from older Zapier-style tools — it’s trying to be the execution layer underneath agents, not just a connector between two SaaS products.

Credits are the unit that matters for pricing, and they burn faster the more steps and AI calls a flow makes, so the free tier’s 100 credits a day is enough to prove a workflow works but not to run it in production. Plus is the realistic entry point for a small business; the jump to Team brings SSO, which matters the moment more than a couple of people touch the account.

Against Make and n8n, the honest comparison is maturity versus openness: Make has more polish and a bigger app library, n8n has a larger self-hosted community, and Activepieces is newer but growing fastest on GitHub, which tends to mean faster iteration on the pieces you’ll actually need next year.

Alternatives

Get Activepieces

Buying the tool is not the hard part.

The ebook is about what to hand over, in what order, and how you find out when it goes wrong. One payment, no subscription.

Join the daily email.

Curated AI news for small businesses and their teams. Free, daily. Sent at 7pm SGT. Unsubscribe anytime.