Mindket
English
English简体中文
Sign inSign up
StoreMCPSkillsCalculatorLearnPriceBlog
Discover
Research308Writing59Business95Legal13Finance & accounting33Marketing & SEO172Developer1,379Productivity382Education81Companions14
Help and support
Language
English简体中文
IntermediateFree

Dify tutorial: build AI apps and workflows without code

Mindket editors · Updated · 4 lessons · 39 min

A black banner titled "Dify" over the line "Build AI apps and workflows without code".

This free Dify course has four lessons. You start Dify on Cloud or with Docker Compose and build a workflow and a chatflow. Then you add a knowledge base with chunking and retrieval settings. Last, you build an agent, publish it as a web app, an embed or an API and read its logs. Every step follows the Dify README and docs.

What you'll be able to do

  • Run Dify on Cloud or on your own machine with Docker Compose, and know what its licence asks.
  • Build a workflow with branches, learn the Code and HTTP Request nodes, then test and debug it.
  • Create a knowledge base, choose its chunk and index settings and connect it to a chatflow.
  • Build an agent app with tools, publish it as a web app, an embed or an API and read its logs.
All courses

Who it's for

For founders, analysts, operations staff and developers new to LLM apps who want to build an AI app or automation without writing a backend. You need a browser for Dify Cloud, or Docker and at least 4 GiB of RAM for a self-hosted install, plus a model provider unless you use the credits a new Cloud account includes.

Lessons

  1. 1Dify tutorial: what it is, how to run it and your first appA Dify tutorial for beginners: what Dify is, how to start on Dify Cloud or with Docker Compose, what its licence asks, and a first chatbot.9 min
  2. 2Dify workflow tutorial: build, run and debug a workflowA Dify workflow tutorial: Workflow or Chatflow, the main nodes, a content generator built from the docs, running and debugging it, and publishing versions.9 min
  3. 3Dify knowledge base tutorial: chunking, indexing and retrievalA Dify knowledge base tutorial: upload files, choose chunk mode and index method, set retrieval, test it and connect it to a chatflow.10 min
  4. 4Dify agent tutorial: add tools, publish and call it from codeA Dify agent tutorial: write the prompt, add tools and a knowledge base, publish as a web app, embed or MCP server, call the API and read the logs.11 min

Frequently asked questions

Is this Dify course free?

Yes. Every lesson is free to read. Dify Cloud has its own plans, and a model provider may bill for the calls your apps make.

Do I need to code to learn Dify?

Not much. The lessons mostly use menus and settings. The few snippets of code, a short Python function, a Jinja template, an iframe and a curl call, are copied from the docs.

Should I use Dify Cloud or host it myself?

Cloud needs nothing installed. A self-hosted Docker Compose install keeps apps and data on a machine you control, and lesson 1 lists what it needs and what the licence asks.

Which version of Dify do the lessons follow?

The docs as they stood on 6 November 2026, when the latest release on GitHub was 1.19.0. Dify changes quickly, so lesson 1 shows how to match a tutorial to the version you run.

What should I learn after this course?

The next course, browser agents, covers AI that clicks, types and browses for you. It starts with what browser agents are.

Share this course

See how others chain steps

Mindket lists n8n automation workflows you can open and read before you use them. They show the same step-by-step thinking a Dify workflow needs.

Browse automations

All courses · All articles