Activepieces logo

Activepieces

AI-enabled automation platform for building chat-driven agents and visual workflows with connected apps.

Activepieces screenshot
Pricing:Free planPaid plans from $20/month

What is Activepieces?

Try Activepieces

Activepieces is an AI-enabled automation platform where you set a goal and an AI agent carries out steps across the apps, flows, and files you allow. You can require human approvals for any step that matters, and paused runs keep their place for up to 30 days. The platform works with your chosen AI provider, supports custom tools, and can be used on Activepieces Cloud or your own instance (including air-gapped options).

What does Activepieces do for my business processes?

Activepieces lets you set a goal for an AI agent and choose which apps, flows, and files it can use. The agent decides the steps and runs them, pausing for your approval before any action you want a person to review first.

Can non-developers build AI-agent automations in Activepieces?

Yes. You can give your agent a goal in plain terms, start free, pick the tools it can use, and approve anything that matters before it runs—then adjust as needed without needing to hand-code every step.

How does Activepieces handle human approvals during an agent run?

You can require approval before the agent acts on steps that matter. When a run is paused for approval, it stays paused for up to 30 days and then can resume from its checkpoint once someone signs off.

Can I use my own AI model and keys with Activepieces?

Yes. An admin sets up the AI provider once (including OpenAI, Anthropic, Google Gemini, Azure OpenAI, or AWS Bedrock), and Activepieces also supports using compatible endpoints so you can keep control of which model you use.

What can an agent do with my existing automations and data?

An agent can use actions from the apps you connect, your own automations, MCP server tools, and files you upload for it to read. You can also set Locked Inputs so the agent can’t change certain values during the run.

Does Activepieces support custom integrations and external MCP servers?

Yes. You can write your own integration in TypeScript, and you can also let the agent use your own MCP servers (including the MCP server live endpoint and tools you expose through it).

What kinds of workflows can I build beyond simple automations?

Activepieces supports advanced workflow logic such as branching, retries, waitpoints, subflows, and error handling, plus human approvals when required. You can also add TypeScript inline for the hardest steps while keeping the rest visual.

Does Activepieces support self-hosting or air-gapped deployment?

Yes. You can run Activepieces as Activepieces Cloud, deploy it on your own infrastructure (Docker or Kubernetes), or use a fully air-gapped option with no outbound network, while keeping enterprise features like projects, SSO, roles, audit logs, and your own secret manager.

Is there a way to use AI providers without locking into one service?

Activepieces supports bringing your own keys and models, with multiple AI providers available (including Anthropic, OpenAI, AWS Bedrock, and Google Gemini) so you can choose what you use.

Last updated
Nov 18, 2026
Date listed
Oct 30, 2026
Pricing checked
Oct 30, 2026