langgenius/dify
Dify is an open-source development platform created for orchestrating large language model applications, agentic workflows, and retrieval-augmented generation pipelines within a visual workspace. The system provides a drag-and-drop workflow canvas, an integrated prompt IDE, automated document extraction for formats including PDF and PPT, and isolated autonomous agent sandboxes for tool execution. It also integrates model providers and observability tools such as Opik, Langfuse, and Arize Phoenix, backed by backend-as-a-service APIs for application integration. Users can deploy Dify through cloud services or self-hosted server environments. The GitHub snapshot shows the repository is not archived.
Self-hosted deployment requires minimum hardware of 2 CPU cores and 4 GiB of RAM, alongside prerequisites of Docker with an active daemon, Docker Compose version 2.35.0 or later, and OpenSSL. Verification confirms the official setup sequence starting from the repository root: change directory using cd docker, copy configuration with cp .env.example .env, generate a secret key using openssl rand -hex 32, set DIFY_AGENT_LOCAL_SANDBOX_AUTH_TOKEN, and launch services via docker compose up -d before visiting http://localhost/install. GitHub records NOASSERTION; the README names the Dify Open Source License, based on Apache 2.0 with additional conditions. Read its license file for the full terms. For another visual agent-workflow platform, compare Langflow; features and licensing differ.
Data as of : 160291 stars · 25345 forks.