Dify is an open-source platform for building LLM applications that lets teams design AI agents, chatbots, RAG pipelines and automated workflows on a visual canvas and ship them as APIs or web apps.
Dify combines the pieces most AI apps need, including model management, prompt orchestration, retrieval over your own documents, tool use and observability, so teams don't have to stitch together separate libraries. It can be self-hosted with Docker or used as a managed cloud service.
Key features
- Visual workflow and chatflow builder with branching, loops and code nodes
- Agent builder with tool calling and agent routing
- Knowledge bases for retrieval-augmented generation (RAG) over your documents
- Support for a wide range of commercial and open-source model providers
- Plugin marketplace and MCP integration
- Publish apps as web apps or call them through APIs
- Logs, annotations and monitoring for production apps
- Self-hosting with Docker Compose or Kubernetes
Who it's for
Dify serves developers, AI engineers and product teams building internal copilots, customer support bots and automated AI workflows, as well as enterprises that need to self-host for data control.
Common use cases
- Building a customer support chatbot over company docs
- Automating document processing and content workflows
- Prototyping and deploying AI agents with tools
- Running a private, self-hosted AI platform
Pricing
Dify's Community edition is free to self-host. Dify Cloud plans, per its pricing page:
- Sandbox: free, with 200 message credits for trying core features.
- Professional: $590 per workspace per year, with 5,000 message credits per month.
- Team: $1,590 per workspace per year, with 10,000 message credits per month.
- Enterprise: custom pricing.
Alternatives
- n8n: workflow automation platform with AI nodes
- RAGFlow: open-source RAG engine with deep document parsing
- Flowise: open-source drag-and-drop LLM app builder
FAQ
Is Dify free?
Yes. Dify's Community edition is free and open source to self-host, and Dify Cloud has a free Sandbox plan.
What is Dify used for?
Dify is used to build and run LLM apps such as chatbots, AI agents, RAG knowledge assistants and automated AI workflows.
Is Dify open source?
Yes. Dify's source code is on GitHub and can be self-hosted with Docker.
Dify vs LangChain: which should I use?
Dify is a visual platform for shipping AI apps quickly, while LangChain is a code library for building custom pipelines. Teams that prefer low-code usually choose Dify.
Dify vs n8n: what's the difference?
Dify is focused on LLM apps, RAG and agents, while n8n is a general workflow automation tool with AI capabilities.
Can Dify use local models?
Yes. Dify supports self-hosted models through providers such as Ollama alongside commercial APIs.