About this project
Dify is an open-source LLM application development platform designed to streamline the process from prototyping to production. It provides a collaborative workspace with a variety of core capabilities:
- **Workflow Orchestration**: A visual canvas for building and testing complex AI workflows.
- **Model Management**: Integration with numerous proprietary and open-source LLMs (including GPT, Mistral, and Llama3) via various inference providers.
- **RAG Pipeline**: Tools for document ingestion and retrieval, with support for extracting text from PDFs, PPTs, and other common formats.
- **Agent Capabilities**: Support for agents based on LLM Function Calling or ReAct, including over 50 built-in tools such as Google Search, DALL·E, and WolframAlpha.
- **Prompt IDE**: An interface for crafting prompts and comparing model performance.
- **LLMOps**: Features for monitoring application logs, analyzing performance, and iteratively improving prompts and datasets.
- **Backend-as-a-Service**: API access to all platform offerings for integration into existing business logic.
Dify can be deployed via Dify Cloud or self-hosted using Docker Compose.