# Higress

> Apache-2.0 AI and API gateway on Envoy and Istio, with Wasm plugins, model proxying and MCP hosting

**Category:** AI Agents  
**Pricing:** open-source  
**URL:** https://higress.ai  
**GitHub:** https://github.com/higress-group/higress  
**Added:** 2026-10-11

---

## The Hook

Higress puts model-provider proxying, API routing and MCP server hosting behind one Envoy-based gateway, so AI traffic gets the same rate limits, auth and observability as the rest of your APIs.

## What's Good

Apache-2.0 (higress-group/higress), release v2.2.5. Built on Envoy and Istio, extended with Wasm plugins you can write in Go, Rust or JS. Proxies many model providers behind one endpoint with token rate limiting, fallback and caching, and can host MCP servers or turn existing REST APIs into MCP tools. Runs on Kubernetes or as a standalone Docker deployment. ★~9.5k.

## The Catch

It is a full Envoy/Istio gateway, much heavier to run than a single-binary model proxy. Default images pull from an Alibaba Cloud registry in Hangzhou, so outside China switch to the US or Southeast Asia mirror if pulls time out. Much of the documentation and the plugin catalog live on higress.cn. Vs agentgateway (a lighter, purpose-built gateway for agent and MCP traffic) and litellm (a Python model proxy you can start in minutes): Higress fits when you want AI traffic inside a production API gateway you already run at scale.

## Verdict

A production-grade AI gateway if you can run Envoy and Istio — for a quick model proxy, start with something smaller.