Logo Dify

Artificial Intelligence

Dify

Build LLM applications with drag-and-drop — chatbots, agents, RAG pipelines — no backend required.

Official documentation Source code

About Dify

Dify is a source-available LLM application development platform that lets you build chatbots, autonomous agents and RAG pipelines without writing any backend code. Its visual interface lets you connect data sources, define conversational flows and deploy APIs in just a few minutes.

Self-hosted on a VPS, Dify gives you full control over your data, your prompts and your models. You can connect any LLM (OpenAI, Anthropic, Mistral, local Ollama models) and build professional AI applications without depending on third-party cloud infrastructure.

What gets installed

Ready when the VPS starts

Compatible OS
Ubuntu 24.04
Application
Dify
Tech stack
DockerPythonPostgreSQLRedisWeaviate
Minimum RAM
4 GB (8 GB recommended)
Minimum CPU
2 vCPU
Default port
80
Access
root over SSH

Key features

  • Visual studio to build LLM workflows without code
  • Native RAG: connect vector databases, PDFs, websites
  • Multi-model support: OpenAI, Claude, Gemini, Mistral, Ollama
  • One-click REST API deployment
  • LLM call monitoring and cost management
  • User and workspace management

When to use this solution?

Business chatbot

Build an intelligent assistant connected to your internal data — documentation, FAQs, CRM — without a single line of backend code.

Document processing pipeline

Automate the analysis, classification and summarization of documents by building a visual workflow.

Rapid AI prototype

Test your LLM application idea in a few hours before investing in custom development.

Guide

Deploy Dify on your VPS

  1. Order the VPS

    A VPS with at least 4 GB of RAM (8 GB recommended), running Ubuntu. Docker and Docker Compose are installed by the recipe: there is nothing for you to prepare.

  2. Install from the Marketplace

    ServOrbit clones the official Dify repository, writes its .env file and starts the stack for you. There is no repository to clone and no file to write.

  3. Sign in for the first time

    Open /install IMMEDIATELY after the installation: Dify asks you to create the administrator account yourself (e-mail + password). Until that account exists the page is open to anyone, and the first visitor to reach it becomes the administrator of your instance.

  4. What the recipe starts

    Dify runs with PostgreSQL, Redis and Weaviate (vector database). All of its published ports are bound to the VPS loopback — its web server listens on host port 18080 — so nothing is reachable in the clear on the public IP.

  5. Expose over HTTPS

    Attach your domain to the app from your ServOrbit client area: the nginx vhost and the TLS certificate are set up for you. Without a domain there is no public access: open an SSH tunnel with ssh -L 18080:127.0.0.1:18080 root@<vps-ip>, then browse http://localhost:18080.

  6. Connect your LLMs

    In the settings, add your API keys (OpenAI, Anthropic…) or the URL of your Ollama instance. Create your first application from the studio.

Frequently asked questions

Questions about Dify

← Back to Artificial IntelligenceAll applications
Dify is a source-available platform that lets you build LLM applications (chatbots, agents, RAG pipelines) through a visual interface, without coding a backend.

Embed the deploy button

Maintaining a project that uses Dify? This button lets your readers deploy it on a VPS in one click, without reading Docker documentation.

Deploy Dify on ServOrbit
Markdown
[![Deploy Dify on ServOrbit](https://servorbit.com/brand/deploy/button.svg)](https://servorbit.com/vps-cloud?template=dify&utm_source=deploy-badge&utm_medium=referral&utm_campaign=dify)
HTML
<a href="https://servorbit.com/vps-cloud?template=dify&utm_source=deploy-badge&utm_medium=referral&utm_campaign=dify"><img src="https://servorbit.com/brand/deploy/button.svg" alt="Deploy Dify on ServOrbit" height="40"></a>

The button points to a VPS order with the template preselected. The image is served from servorbit.com — nothing to host on your side.

Complementary solutions

Artificial Intelligence

Open WebUI

Web interface to interact with your local or remote LLMs. Your data stays on your infrastructure — no third party involved.

Automation & Workflows

Flowise

Build LLM automation pipelines with drag-and-drop. Visual LangChain chains, connected to your existing tools.

Artificial Intelligence

OpenClaw

Your open-source personal AI assistant, reachable from WhatsApp, Telegram, Discord and 50+ messaging apps. Self-hosted on your VPS — your conversations stay with you.

Need help?

Browse our help center and FAQ, or reach our team — callback, WhatsApp or email.

Message us on WhatsAppopens in a new tab
or write to us