Deploy Dify AI on Northflank

Dify is an open source platform for building, deploying, and operating AI applications such as chatbots, AI agents, workflows, and internal tools. It’s widely used by teams that want full control over models, data, plugins, and infrastructure while running production-ready AI systems.

With Northflank, you can deploy Dify in minutes using the Dify AI stack template. This prebuilt setup provisions all required services, storage, networking, and secrets automatically, so you can focus on building AI applications instead of managing complex infrastructure.

If you’re new to the platform, Dify is a self-hosted AI application development platform that combines prompt management, workflows, vector search, plugins, and runtime execution into a single system.

Dify lets you build AI-powered apps using APIs or a web UI, store application data in PostgreSQL, manage background tasks with Redis and workers, execute code securely in a sandbox, and power semantic search with a vector database. Many teams self-host Dify to run internal tools, customer-facing AI apps, and production AI workflows without relying on third-party SaaS limits.

The Dify AI stack template provisions everything needed for a production-ready Dify environment.

It includes:

• PostgreSQL database for core Dify data and application state
• Redis addon for task queues, caching, and background job processing
• MinIO object storage for files, assets, and plugin storage
• Dedicated PostgreSQL database for plugin data
• Weaviate vector database for embeddings and semantic search
• Dify API service for application logic and AI orchestration
• Dify Web service for the console and frontend UI
• Dify Worker and Worker Beat services for background tasks and scheduling
• Dify Sandbox service for secure code execution
• Dify Plugin Daemon for managing and running plugins
• Persistent volumes for Weaviate and the sandbox runtime
• Secret groups for securely managing credentials, API keys, and internal service configuration

This architecture separates storage, compute, workers, and vector search, ensuring the platform remains responsive and scalable under load.

1. Create an account on Northflank
2. Click Deploy Dify AI now
3. Click Deploy stack to save and run the Dify template
4. Wait for all services to finish deploying
5. Open the public Dify Web URL to access the Dify console
6. Start building AI applications, workflows, and plugins

No manual configuration is required. All internal connections, credentials, and service URLs are set automatically.

This stack template gives you a complete, production-ready Dify setup:

• Run self-hosted AI applications with full control over data and infrastructure
• Store structured data securely in PostgreSQL with TLS enabled
• Handle background tasks and scheduling with Redis and dedicated workers
• Power semantic search and embeddings using Weaviate
• Store files and plugin assets using S3-compatible object storage via MinIO
• Execute code safely in an isolated sandbox environment
• Manage plugins with a dedicated plugin daemon and database
• Securely manage secrets, API keys, and internal service credentials
• Scale API, workers, and frontend services independently

It follows production best practices and is suitable for both internal tools and customer-facing AI applications.

• Dify Web – Hosts the web console and frontend UI for managing apps and workflows
• Dify API – Core backend service handling requests, AI logic, and orchestration
• Dify Worker & Beat – Process background jobs, async tasks, and scheduled workflows
• PostgreSQL (DifyDB) – Stores core application data, users, configurations, and state
• PostgreSQL (PluginDB) – Isolated database for plugin-related data
• Redis – Powers queues, caching, and task coordination
• Weaviate – Vector database used for embeddings and semantic retrieval
• MinIO – Object storage for files, datasets, and plugin assets
• Dify Sandbox – Secure environment for executing code and tool logic
• Plugin Daemon – Manages plugin lifecycle and remote plugin installation
• Secret Groups – Automatically inject database credentials, API keys, endpoints, and runtime configuration
• Load Balancers – Distribute traffic efficiently to keep services responsive

All services communicate over private networking, with only the Dify Web and API exposed publicly.

Deploying Dify on Northflank is the fastest way to run a scalable, secure, self-hosted AI application platform in production.

With this stack template, you get a fully wired Dify environment including databases, vector search, object storage, workers, plugins, and secure runtime execution. You can start building AI-powered products immediately without worrying about infrastructure, networking, or secret management.