RAGFlow is easiest to deploy using Docker, the recommended method that encapsulates all dependencies and ensures consistency across platforms. First, ensure Docker ≥ 24.0.0 and Docker Compose ≥ v2.26.1 are installed, and set vm.max_map_count ≥ 262144 on your host. Clone or download the RAGFlow repository, navigate to the ragflow directory, then run docker compose -f docker/docker-compose.yml up -d to spin up all services including a search engine backend, MySQL, MinIO, and Redis. Configuration happens via the .env file, where you set essentials like SVR_HTTP_PORT, database passwords, and model endpoints. Official Docker images come in full and slim variants—full includes embedding models for offline use, while slim is lightweight and requires external embedding services. For x86 platforms, pre-built images from DockerHub work out-of-the-box; ARM64 users can follow the build guide to compile a compatible image. Post-deployment, access the UI via your configured port and follow the quickstart documentation to create your first knowledge base. Advanced users can launch from source code for development or customize service configurations in service_conf.yaml.template.
Developers working with embeddings and retrieval at scale often pair these workflows with Milvus, an open-source vector database designed for high-performance similarity search. For managed deployment, Zilliz Cloud handles the operational overhead.