Repolode
Submit
Kong

Kong/kong

Kong is a cloud-native, platform-agnostic API, LLM and MCP gateway, extensible through plugins, for routing and securing service traffic.

Open-source alternative to:ApigeeAWS API Gateway

View on GitHubWebsite
Stars
44,288
Forks
5,235
Open issues
220
Last commit
2 days ago

What is it?

Kong is a gateway that sits in front of microservices and conventional APIs, handling proxying, routing, load balancing, health checking and authentication. Its functionality is extended through plugins, and it runs natively on Kubernetes via its official Ingress Controller.

Kong also covers AI traffic. It offers a universal LLM API for routing requests across providers such as OpenAI, Anthropic, Google Gemini, AWS Bedrock, Azure AI, Mistral and Hugging Face, along with MCP traffic governance, security and observability, and features like semantic caching and semantic routing.

Configuration can be done through a RESTful Admin API or declarative files. Kong supports DB-less declarative deployment as well as hybrid deployments that separate the control plane from the data plane. Plugins can be written in Lua, Go or JavaScript.

Key features

  • 01Routing, load balancing and health checking configurable via Admin API or declarative config.
  • 02Authentication and authorization with JWT, basic auth, OAuth and ACLs.
  • 03Unified LLM API for routing across multiple AI model providers.
  • 04MCP traffic governance, security and observability.
  • 05Plugin system for rate limiting, request/response transformation and logging.
  • 06Native Kubernetes Ingress Controller support.

Use cases

  • Putting a central gateway in front of many microservices to handle auth and rate limiting.
  • Routing and monitoring calls from applications to multiple LLM providers.
  • Running API traffic on Kubernetes clusters with an ingress controller.
  • Managing and securing MCP server traffic for AI agents.

Quick start

  1. 1

    Clone the Docker repository

    git clone https://github.com/Kong/docker-kong
  2. 2

    Go to the compose folder

    cd docker-kong/compose/
  3. 3

    Start the Gateway stack with Postgres

    KONG_DATABASE=postgres docker-compose --profile database up

Things to know

  • The quick start relies on Docker Compose, so Docker must be installed first.
  • Advanced AI gateway features are documented separately and may require a Kong Konnect or commercial setup.

Topics

#api-gateway#reverse-proxy#microservices#kubernetes#llm-gateway#mcp#api-management#ai#ai-gateway#apis

FAQ

What is kong?
Kong is a cloud-native, platform-agnostic API, LLM and MCP gateway, extensible through plugins, for routing and securing service traffic. Kong is a gateway that sits in front of microservices and conventional APIs, handling proxying, routing, load balancing, health checking and authentication. Its functionality is extended through plugins, and it runs natively on Kubernetes via its official Ingress Controller.
Is kong free to use? What license does it have?
kong is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is kong still maintained?
Actively maintained. The last commit to kong was on Oct 19, 2026.
What can I use kong for?
Putting a central gateway in front of many microservices to handle auth and rate limiting. Routing and monitoring calls from applications to multiple LLM providers. Running API traffic on Kubernetes clusters with an ingress controller. Managing and securing MCP server traffic for AI agents.
How do I install kong?
Clone the Docker repository: git clone https://github.com/Kong/docker-kong Go to the compose folder: cd docker-kong/compose/ Start the Gateway stack with Postgres: KONG_DATABASE=postgres docker-compose --profile database up
Where can I get kong?
The source code and installation instructions are on GitHub: https://github.com/Kong/kong

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects