Being AI Ready
AI Automation

Activepieces

An open-source, AI-first automation platform priced per active workflow rather than per task or credit, with a self-hostable MIT-licensed Community Edition as an alternative to Zapier or n8n.

Visit website

What it does

Activepieces is an open-source automation platform that positions itself as “AI-first automation for every team,” combining a drag-and-drop workflow builder with genuinely open-source roots — its Community Edition is MIT-licensed and self-hostable, with more than 270 contributors on GitHub. That makes it one of two self-hostable options in this category alongside n8n, appealing to teams that want infrastructure-level control or simply prefer not to depend entirely on a vendor’s cloud.

Its pricing model is also distinct: rather than charging per completed task (Zapier) or per credit consumed (Make, Gumloop), Activepieces charges per active workflow, with unlimited task executions included once a flow is active. For teams running a small number of high-volume automations, this can be considerably cheaper than usage-metered pricing; for teams running many small, low-frequency flows, it can work out the other way. With more than 700 integrations and unlimited MCP (Model Context Protocol) server support on paid tiers, Activepieces also makes it straightforward to expose its automations as tools that AI coding assistants and chat interfaces can call directly, rather than only triggering on a schedule or webhook.

Compared to the two established incumbents, Activepieces is a smaller, newer ecosystem — its documentation and community template library are less extensive than Zapier’s or Make’s, and its native AI-agent orchestration is generally considered less mature than n8n’s. It’s a strong fit for technical teams that specifically want open-source self-hosting or per-flow pricing, and less clearly the right default for a team that just wants the most battle-tested, broadest-catalog automation tool available.

Ideal users

  • Teams that want an open-source, self-hostable alternative to Zapier with a genuinely permissive (MIT-licensed) community edition
  • Organizations running many active workflows that want predictable, per-flow pricing rather than task- or credit-based billing that scales with execution volume
  • Technical teams that want to build AI agents and expose automations as MCP servers for use by AI coding assistants and chat tools

Who should avoid it

  • You want the largest, most mature integration catalog with the most polished pre-built connectors — Zapier and Make have larger, longer-established libraries
  • You're not comfortable running or maintaining self-hosted infrastructure and also don't want per-active-flow cloud pricing to scale with how many separate automations you run
  • You want a fully managed, natural-language AI assistant experience — Lindy is closer to that than Activepieces' node-based builder
  • You need the deepest AI-agent tooling (multi-agent orchestration, RAG) — n8n's AI-agent features are generally considered more mature

Key features

  • 700+ app integrations spanning sales, support, marketing, finance, HR, IT, and engineering use cases
  • AI agent building with enterprise integrations and configurable access controls
  • MIT-licensed, self-hostable Community Edition with 270+ open-source contributors
  • Per-active-flow pricing with unlimited task executions on paid tiers, rather than metering every run
  • Unlimited MCP (Model Context Protocol) servers, letting AI assistants and coding tools trigger Activepieces automations directly
  • Pre-built AI agent templates for lead qualification, meeting summarization, inbox management, and deal-closing tasks

Pros

  • Genuinely open-source (MIT-licensed core) with an active self-hosting community, not just a marketing claim
  • Per-active-flow pricing with unlimited runs can be significantly cheaper than task- or credit-based billing at high execution volume
  • Large integration count (700+) for a tool of its size and age
  • Strong native MCP support makes it easy to expose automations to AI coding assistants and chat tools

Cons

  • Smaller, newer ecosystem than Zapier or Make, with less battle-tested documentation and fewer community templates
  • Self-hosting the Community Edition requires technical setup and maintenance, similar to n8n
  • AI-agent orchestration is less mature than n8n's, based on available comparisons
  • Per-active-flow pricing can work against you if you run many small, low-volume flows rather than a few high-volume ones

Pricing

Free — $5/month per active flow (Standard plan, after the first 10 free active flows)

As of July 2026, Activepieces' pricing has two tiers: Standard, which is free for the first 10 active flows and then $5 per additional active flow per month with unlimited runs, AI agents, and MCP servers; and Ultimate, a custom-priced annual plan adding team/personal projects, piece access controls, custom RBAC, SSO, and audit logs. The Community Edition remains free to self-host under an MIT-based license with core features only. Confirmed directly at activepieces.com/pricing.

Typical workflows

  • A support team self-hosts Activepieces' Community Edition and builds an automation that summarizes new tickets with AI, tags them by urgency, and routes only high-priority tickets to a Slack channel.
  • A developer exposes several Activepieces automations as MCP servers so an AI coding assistant can trigger a deployment workflow or check a build status directly from a chat interface.

Integrations

  • 700+ app integrations (Gmail, Slack, Notion, HubSpot, Salesforce, Shopify, GitHub, and more)
  • OpenAI and other LLM providers for AI agent steps
  • Model Context Protocol (MCP) servers, unlimited on paid plans
  • Custom API connections via HTTP pieces

Privacy & security notes

Because Activepieces can be fully self-hosted under its MIT-licensed Community Edition, it offers strong data control for teams that need automation data to stay on infrastructure they manage. The hosted cloud version processes data on Activepieces' own servers like any SaaS tool, retaining personal data only as needed for service operation and legal compliance and deleting it within 12 months of account termination; its privacy policy states customer data is not used to train or improve AI/ML models, and named subprocessors include Stripe, Google Analytics, Intercom, Mixpanel, and Userback.

Frequently asked questions

Is Activepieces really free and open source?

The Community Edition is MIT-licensed and free to self-host, covering core automation features with an active open-source contributor base. The hosted cloud version uses a per-active-flow pricing model, with 10 flows free and paid tiers beyond that for additional flows, AI agents, and team features.

How does 'per active flow' pricing compare to Zapier's per-task pricing?

Activepieces charges based on how many workflows you keep active, with unlimited executions within each flow, while Zapier charges per completed task regardless of how many flows you have. This means Activepieces can be considerably cheaper if you run a small number of high-volume workflows, but potentially more expensive if you run many separate low-volume flows.

Can Activepieces automations be triggered by AI coding assistants like Claude or Cursor?

Yes, Activepieces supports unlimited MCP (Model Context Protocol) servers, which lets AI coding assistants and chat tools call Activepieces automations directly as tools, rather than only running on a schedule or webhook trigger.

How does Activepieces compare to n8n, the other self-hostable option in this category?

Both are self-hostable and developer-friendly, but n8n has a more mature AI-agent and code-step ecosystem, while Activepieces emphasizes a larger out-of-the-box integration count and simpler per-active-flow pricing. Teams choosing between them should trial both against a real workflow rather than deciding from feature lists alone.

Best alternatives