Connect business apps and build agents, flows and tables with cloud or self-hosted deployment
Activepieces is an automation platform that connects the applications a team already uses and turns repeated work into executable flows. It combines a visual automation foundation with AI agents, chat and tables. A workflow can respond to an app event, process information and carry out actions in connected tools, while an agent can use instructions and tools to handle a less rigid task. The platform supports both managed cloud use and deployment on an organization’s own infrastructure.
Integrations are called pieces, and the catalog includes tools such as Gmail, Slack, Notion, HubSpot and OpenAI. HTTP requests and webhooks provide another route for working with services that expose an API. Custom logic and human approval can be included in the workflow, giving builders a way to combine automated steps with review. Tables provide a place for the structured information used by these processes.
Developers can also create a dedicated integration using the open-source TypeScript piece framework. Paid plans support bringing an organization’s own AI keys, while hosted AI choices use different model tiers. Chat-to-automation is described as a gradual rollout, so its presence should be checked in the actual workspace before planning a process around it.
Organizations can choose cloud hosting or self-hosting. Team and enterprise capabilities include access controls, shared connections and administrative features, with more advanced governance offered on higher plans. The free self-hosted Community Edition focuses on the automation core; it should not be confused with the complete hosted product because agents, chat and several administrative features sit outside its scope.
Team adds shared projects, SSO and standard roles. Ultimate extends governance with capabilities such as SCIM provisioning, custom permissions, audit logs, secret managers and event streaming. Git Sync and releases support organizations that need a controlled process for moving automation changes between environments. These administrative differences matter when automations become shared operational infrastructure rather than one person’s convenience tool.
The hosted free plan includes a daily credit allowance. A flow run and AI activity have different credit costs, so an agent-heavy workflow can consume more than a simple automation. Paid tiers add larger allowances, more users and additional controls. Suitable starting projects include routing messages, updating CRM records or preparing summaries from connected data. Builders can test a small flow first, inspect its result and then expand it into a broader process once the connections and conditions behave as intended.
