Kong Gateway
A self-hosted API and AI gateway that routes LLM and MCP traffic, supports Kubernetes, and extends traffic controls through plugins.
Kong Gateway gives platform teams a shared place to manage API, LLM and MCP requests. It helps teams apply access and traffic policies across services and AI applications on infrastructure they control.
- API traffic controls: Route requests, balance traffic across services, check service health, and apply authentication, authorization and rate limits.
- LLM routing: A common LLM API connects applications to providers including OpenAI, Anthropic, GCP Gemini, AWS Bedrock, Azure AI and Mistral.
- AI request policies: AI capabilities include semantic routing and caching, security checks, and observability for model traffic.
- MCP management: Govern and inspect MCP traffic, apply security controls, and generate MCP interfaces from RESTful APIs.
- Plugin extensions: Add request and response changes, logging, monitoring, or custom gateway behavior through plugins.
The open-source gateway uses Apache 2.0 and can run on your own infrastructure, including Kubernetes through its official Ingress Controller. It supports database-free operation and deployments that separate central policy management from the components handling requests. Calls routed to cloud model providers still go to those providers.
Kong offers commercial subscriptions with additional gateway functionality. Kong Konnect provides a managed cloud control plane, analytics, a service catalog and developer portals.
Logotyper
