Repolode
Submit
langgenius

langgenius/dify

Open-source LLM app development platform combining visual agentic workflows, RAG pipelines, model management and observability in one workspace.

Open-source alternative to:FlowiseLangChain Studio

View on GitHubWebsite
Stars
158,940
Forks
25,121
Open issues
1,146
Last commit
6 hours ago

What is it?

Dify is an open-source platform for building LLM applications. It brings together a visual workflow canvas, a prompt IDE for comparing model outputs, a RAG pipeline that handles document ingestion and retrieval, and agent capabilities, so teams can move an app from prototype to production inside one tool.

Model support covers hundreds of proprietary and open-source LLMs from many inference providers, including GPT, Mistral, and Llama 3, as well as any OpenAI API-compatible endpoint. Agents can run commands, handle files in their own sandbox, and call tools from the Dify Marketplace, MCP servers, or custom APIs. Application logs and performance are monitored over time, with integrations for Opik, Langfuse, and Arize Phoenix.

Every application built in Dify can be exposed through backend APIs, making it possible to call Dify apps from existing business systems. Dify can be used through its cloud service or self-hosted with Docker Compose.

Key features

  • 01Visual workflow canvas for building and testing AI workflows
  • 02Support for hundreds of LLMs from many providers and any OpenAI API-compatible model
  • 03Prompt IDE for crafting prompts and comparing model performance
  • 04RAG pipeline with text extraction from PDFs, PPTs, and other document formats
  • 05Agents with a sandbox, skills, and tools via Marketplace, MCP servers, or custom APIs
  • 06Backend APIs for every app, plus logs and performance monitoring

Use cases

  • Prototyping a chatbot grounded in internal documents with RAG
  • Building multi-step agent workflows that call external tools and APIs
  • Letting teams compare models and prompts before choosing one for production
  • Exposing LLM-powered features to existing products through backend APIs

Quick start

  1. 1

    Go to the dify project folder

    cd dify
  2. 2

    Go to the docker folder

    cd docker
  3. 3

    Copy the example environment file

    cp .env.example .env
  4. 4

    Start Dify with Docker Compose

    docker compose up -d

Things to know

  • The license is the Dify Open Source License, based on Apache 2.0 with additional conditions, so review it before commercial use.
  • Self-hosting requires Docker and Docker Compose v2.24.0 or later, plus at least 2 CPU cores and 4 GiB RAM.
  • Seeking help beyond the README's FAQ means relying on community channels.

Topics

#llm#agent#rag#workflow#low-code#prompt-engineering#openai-api#self-hosted#agentic-ai#agentic-framework

FAQ

What is dify?
Open-source LLM app development platform combining visual agentic workflows, RAG pipelines, model management and observability in one workspace. Dify is an open-source platform for building LLM applications. It brings together a visual workflow canvas, a prompt IDE for comparing model outputs, a RAG pipeline that handles document ingestion and retrieval, and agent capabilities, so teams can move an app from prototype to production inside one tool.
Is dify free to use? What license does it have?
dify is open source under the Other license. Check the license terms before using it in commercial products.
Is dify still maintained?
Actively maintained. The last commit to dify was on Oct 20, 2026.
What can I use dify for?
Prototyping a chatbot grounded in internal documents with RAG Building multi-step agent workflows that call external tools and APIs Letting teams compare models and prompts before choosing one for production Exposing LLM-powered features to existing products through backend APIs
How do I install dify?
Go to the dify project folder: cd dify Go to the docker folder: cd docker Copy the example environment file: cp .env.example .env Start Dify with Docker Compose: docker compose up -d
Where can I get dify?
The source code and installation instructions are on GitHub: https://github.com/langgenius/dify

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects