Features and Specifications

For those already familiar with LLM application tech stacks, this document serves as a shortcut to understand Dify's unique advantages

We adopt transparent policies around product specifications to ensure decisions are made based on complete understanding. Such transparency not only benefits your technical selection, but also promotes deeper comprehension within the community for active contributions.

Project Basics

Established	March 2023
Open Source License	Apache License 2.0 with commercial licensing
Official R&D Team	Over 15 full-time employees
Community Contributors	Over 290 people(As of Q2 2024)
Backend Technology	Python/Flask/PostgreSQL
Frontend Technology	Next.js
Codebase Size	Over 130,000 lines
Release Frequency	Average once per week

Technical Features

LLM Inference Engines	Dify Runtime (LangChain removed since v0.4)
Commercial Models Supported	10+, including OpenAI and Anthropic Onboard new mainstream models within 48 hours
MaaS Vendor Supported	7, Hugging Face, Replicate, AWS Bedrock, NVIDIA, GroqCloud, together.ai,, OpenRouter
Local Model Inference Runtimes Supported	6, Xoribits (recommended), OpenLLM, LocalAI, ChatGLM,Ollama, NVIDIA TIS
OpenAI Interface Standard Model Integration Supported	∞
Multimodal Capabilities	ASR Models Rich-text models up to GPT-4o specs
Built-in App Types	Text generation, Chatbot, Agent, Workflow, Chatflow
Prompt-as-a-Service Orchestration	Visual orchestration interface widely praised, modify Prompts and preview effects in one place. Orchestration Modes Simple orchestration Assistant orchestration Flow orchestration Prompt Variable Types String Radio enum External API File (Q3 2024)
Agentic Workflow Features	Industry-leading visual workflow orchestration interface, live-editing node debugging, modular DSL, and native code runtime, designed for building more complex, reliable, and stable LLM applications. Supported Nodes LLM Knowledge Retrieval Question Classifier IF/ELSE CODE Template HTTP Request Tool
RAG Features	Industry-first visual knowledge base management interface, supporting snippet previews and recall testing. Indexing Methods Keywords Text vectors LLM-assisted question-snippet model Retrieval Methods Keywords Text similarity matching Hybrid Search N choose 1（Legacy） Multi-path retrieval Recall Optimization Rerank models
ETL Capabilities	Automated cleaning for TXT, Markdown, PDF, HTML, DOC, CSV formats. Unstructured service enables maximum support. Sync Notion docs as knowledge bases. Sync Webpages as knowledge bases.
Vector Databases Supported	Qdrant (recommended), Weaviate，Zilliz/Milvus, Pgvector, Pgvector-rs，Chroma, OpenSearch, TiDB, Tencent Cloud VectorDB, Oracle, Relyt, Analyticdb, Couchbase, OceanBase, Tablestore, Lindorm, OpenGauss
Agent Technologies	ReAct, Function Call. Tooling Support Invoke OpenAI Plugin standard tools Directly load OpenAPI Specification APIs as tools Built-in Tools 40+ tools(As of Q2 2024)
Logging	Supported, annotations based on logs
Annotation Reply	Based on human-annotated Q&As, used for similarity-based replies. Exportable as data format for model fine-tuning.
Content Moderation	OpenAI Moderation or external APIs
Team Collaboration	Workspaces, multi-member management
API Specs	RESTful, most features covered
Deployment Methods	Docker, Helm