langgenius/dify

What is Dify?

Dify (by LangGenius) is an open‑source platform for building applications that use large language models (LLMs). It provides a web‑based UI where you can design AI workflows, connect retrieval‑augmented generation (RAG) pipelines, and create autonomous agents without writing code. The platform also includes tools for model management, prompt editing, observability (logs, performance dashboards), and a backend‑as‑a‑service API so the apps you build can be integrated into other systems.

Who might use it?

• Product teams that want to prototype an LLM‑powered chatbot, document‑search tool, or any AI‑enhanced feature quickly.
• Developers who prefer a visual canvas for wiring together LLM calls, data retrieval, and tool‑use, but still need the ability to export or call the workflow via API.
• Enterprises looking for a self‑hosted solution with SSO, RBAC, and support options (via the Dify Enterprise offering).
• Researchers or hobbyists who want to experiment with many model providers (OpenAI, Mistral, Llama 3, etc.) and compare prompt performance.

Core features (as described in the README)

Feature	What it does
Workflow canvas	Drag‑and‑drop UI to chain together LLM calls, retrieval steps, and custom tools.
Model support	Connect to dozens of inference providers and self‑hosted models; supports GPT, Mistral, Llama 3 and any OpenAI‑compatible API.
Prompt IDE	Write, test, and compare prompts; add extras like text‑to‑speech.
RAG pipeline	Ingest PDFs, PPTs, etc., extract text, index it, and retrieve relevant chunks during a chat.
Agent sandbox	Define autonomous agents that can run commands, install software, and manipulate files; extend them with tools from the Dify Marketplace or custom APIs.
LLMOps / observability	Built‑in logging, performance dashboards, and integrations with Opik, Langfuse, Arize Phoenix for monitoring.
Backend‑as‑a‑Service	All UI actions are also exposed via RESTful APIs for programmatic use.
Deployment options	Cloud‑hosted SaaS, self‑hosted Community Edition (Docker‑Compose), and Enterprise self‑hosted with SSO/RBAC.

How to get started

1. Prerequisites – at least a 2‑core CPU and 4 GiB RAM.
2. Docker‑Compose – clone the repo, copy .env.example to .env, then run docker compose up -d inside the docker folder.
3. Open http://localhost/install in a browser and follow the on‑screen setup wizard.
4. For deeper customization, the README points to an “advanced setup” guide and a source‑code deployment guide.

Community & support

• Discord, GitHub Discussions, Issues, and a Twitter/X account for announcements.
• Dify Cloud offers a hosted version for quick trials.
• Enterprise contact form for organizations needing dedicated support.

License

The code is released under the Dify Open Source License, which is based on Apache 2.0 with additional conditions (see LICENSE).

---

All information above is taken directly from the repository’s README; no external assumptions have been added.