Kong/kong

What it solves

Kong Gateway provides a centralized layer for managing and orchestrating traffic for microservices, conventional APIs, and AI-driven workloads. It eliminates the need for engineering teams to implement repetitive infrastructure tasks—like authentication, routing, and load balancing—individually for every service.

How it works

Kong acts as a cloud-native proxy that sits between clients and services. It uses a plugin-based architecture to extend its functionality, allowing users to add capabilities like rate limiting or logging. For AI workloads, it provides a universal LLM API that can route traffic across different providers (such as OpenAI, Anthropic, and Azure AI) and manages Model Context Protocol (MCP) traffic for agentic workflows.

Who it’s for

It is designed for platform engineers and developers who need to manage scalable API traffic, orchestrate LLM-based applications, or secure and monitor MCP traffic in a Kubernetes or bare-metal environment.

Highlights

• Universal LLM API: Route requests across multiple LLM providers through a single interface.
• MCP Governance: Provides security, observability, and autogeneration of MCP from RESTful APIs.
• AI-Specific Tooling: Includes semantic security, semantic routing, and AI caching.
• Extensible Architecture: Over 60 AI features and a vast plugin hub supporting Lua, Go, and JavaScript.