Helm chart for deploying the Harbor container registry onto a Kubernetes cluster.
BorisPolonsky/dify-helm
A Helm chart for deploying the Dify LLM app platform on Kubernetes, covering its API, web, worker, sandbox and plugin components.
- Stars
- 676
- Forks
- 189
- Open issues
- 5
- Last commit
- 8 days ago
What is it?
dify-helm is a Helm chart that deploys langgenius/dify, an LLM-based chatbot and application platform, onto a Kubernetes cluster. A single helm install command sets up the Dify API, web frontend, Celery worker and beat, sandbox for code execution, plugin daemon, agent backend, and an Nginx proxy that routes traffic between them.
The chart can run with bundled dependencies or point to external services. The README lists support for external PostgreSQL or MySQL, Redis (standalone or Sentinel), object storage such as S3, Azure Blob, GCS and Alibaba OSS, and a range of vector databases including Weaviate, Qdrant, Milvus and PGVector. It requires Kubernetes 1.23+ and Helm 3.12+, and the chart is published through a Helm repository and Artifact Hub.
Key features
- 01Deploys the full Dify stack, including API, web, worker, sandbox and plugin daemon, with one Helm install.
- 02Includes an Nginx proxy that routes requests to the API, web and plugin services.
- 03Works with external PostgreSQL or MySQL, and Redis in standalone or Sentinel mode.
- 04Supports several object storage backends such as S3, Azure Blob, GCS and Alibaba OSS.
- 05Supports a range of external vector databases, including Weaviate, Qdrant, Milvus and PGVector.
Use cases
- Running a self-hosted Dify instance on an existing Kubernetes cluster.
- Deploying Dify with managed PostgreSQL, Redis and object storage instead of in-cluster databases.
- Standing up separate Dify environments for staging and production with customized Helm values.
Quick start
- 1
Add the Helm repository
helm repo add dify https://borispolonsky.github.io/dify-helm - 2
Update the local repository index
helm repo update - 3
Install the Dify chart into your cluster
helm install my-release dify/dify
Things to know
- Requires a Kubernetes cluster (1.23+) and Helm 3.12+.
- Deploys many components (API, web, worker, sandbox, plugin daemon), so self-hosting takes real setup effort.
- Customized installs require reading the separate chart README.
In collections
Topics
FAQ
- What is dify-helm?
- A Helm chart for deploying the Dify LLM app platform on Kubernetes, covering its API, web, worker, sandbox and plugin components. dify-helm is a Helm chart that deploys langgenius/dify, an LLM-based chatbot and application platform, onto a Kubernetes cluster. A single helm install command sets up the Dify API, web frontend, Celery worker and beat, sandbox for code execution, plugin daemon, agent backend, and an Nginx proxy that routes traffic between them.
- Is dify-helm free to use? What license does it have?
- dify-helm is open source under the MIT license. Check the license terms before using it in commercial products.
- Is dify-helm still maintained?
- Actively maintained. The last commit to dify-helm was on Oct 15, 2026.
- What can I use dify-helm for?
- Running a self-hosted Dify instance on an existing Kubernetes cluster. Deploying Dify with managed PostgreSQL, Redis and object storage instead of in-cluster databases. Standing up separate Dify environments for staging and production with customized Helm values.
- How do I install dify-helm?
- Add the Helm repository: helm repo add dify https://borispolonsky.github.io/dify-helm Update the local repository index: helm repo update Install the Dify chart into your cluster: helm install my-release dify/dify
- Where can I get dify-helm?
- The source code and installation instructions are on GitHub: https://github.com/BorisPolonsky/dify-helm
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Helm is the package manager for Kubernetes, installing and managing apps packaged as reusable Charts.
Open source cloud native container registry that stores, signs, and scans images and Helm charts, with RBAC and replication.
A declarative GitOps continuous delivery tool that syncs Kubernetes applications to the state defined in Git repositories.
A Python client library for Kubernetes with an API modeled on kubectl, offering sync and async (asyncio/trio) interfaces.
A desktop Kubernetes client built with Tauri and Rust that reads your kubeconfig and derives status the way kubectl does.