Installation
You can install Kong on most Linux distributions and MacOS. We even provide the source so you can compile it yourself.
Install Kong →
5-minute Quickstart
Learn how to start Kong, add a Service, enable plugins, and add consumers in under thirty seconds.
Start using Kong →
DB-less & Declarative Configuration
Learn how to leverage the declarative configuration format for using Kong without a database, using in-memory storage only.
Read the tutorial →Upgrade guide
Already using Kong, and wanting to upgrade? Here's the step-by-step guide.
Read the upgrade guide →Configuration file
Want to further optimize your Kong cluster, database, or configure NGINX? Dive into the configuration.
Start configuring Kong →CLI reference
Want a better understanding of the CLI tool and its options? Browse the detailed command reference.
Use the CLI →Admin API reference
Ready to learn the underlying interface? Browse the Admin API reference to learn how to start making requests.
Explore the interface →Proxy reference
Learn every way to configure Kong to proxy your Services, serve them over SSL or use WebSockets.
Read the Proxy Reference →Load balancing reference
Learn how to setup Kong to load balance traffic through replicas of your upstream services.
Read the Load balancing Reference →Health checks & circuit breakers
Let Kong monitor the availability of your services and adjust its load balancing accordingly.
Learn about health checks and circuit breakers →
Clustering
If you are starting more than one node, you must use clustering to make sure all the nodes belong to the same Kong cluster.
Read the clustering reference →Write your own plugins
Looking for something Kong does not do for you? Easy: write it as a plugin. Learn how to write your own plugins for Kong.
Read the plugin development guide →Go Plugins
You can also write plugins using the Go programming language. Here's how to get started.
Read the guide →Hybrid Mode
Get started with Hybrid Mode, through which you can configure Kong nodes with dedicated roles: you can have Control Plane nodes using a database and Data Plane nodes using DB-less mode.
Read the tutorial →