Start from one repetitive task, let AI do the reading and drafting step, and keep a person on the decisions.
For anyone asking: How do I use AI to automate work in my business?
Each one's way of doing it, and the point where it stops.
An AI node picks a model from a dropdown without a separate provider account; a condition node branches on its label, and RAG storage grounds drafts in your own documents. Where it stops: Cloud-only, no self-hosting; agents are less predictable than a fixed flow, so keep scope narrow and a human in the loop.
The AI Agent node calls tools such as HTTP requests, databases, sub-workflows or an MCP server, and a gated tool waits until a person approves it in Slack, Teams, Gmail or Outlook. Where it stops: Community Edition is free but you run and maintain the server; standalone n8n Agents are in preview.
An AI module in any scenario step summarizes, classifies, drafts or extracts, and a router branches on its answer; AI Agents decide across connected apps on the same canvas. Where it stops: AI steps use more credits than plain data moves, and agent loops drain the allowance faster; AI Agents are in beta.
Copilot drafts a Zap from a plain-English description; inside the Zap an AI step summarizes, classifies or extracts text, and Filter or Paths routes the result to the right app. Where it stops: Free allows only two-step Zaps, so a multi-step AI workflow needs Professional; Agents are not ready for unsupervised complex conditional logic.
An AI app step such as OpenAI takes data from earlier steps into its prompt; free Filter and Formatter steps gate the input and shape the answer before a Router sends it on. Where it stops: Create using AI suggests steps, not a finished workflow; branching is built by hand, and your AI provider bills its usage separately.
Copilot (beta) drafts the trigger and actions from a plain-language request; AI steps then extract fields from inbound text, classify requests by topic and urgency, or score leads. Where it stops: AI is not on Free; Albato does not publicly disclose which model powers its native AI features.
AI Prompt, AI Summarize and AI Image actions take fields from the trigger, such as an email subject, and pass the output to the next action, for example a Slack post. Where it stops: One trigger-to-action chain: no branching, no memory between runs, and the AI does not decide what to do next.
Every verdict rests on the data in our product reports, never on a fee. How we score →