An open source control plane built on Envoy that gives apps one OpenAI-compatible API across hosted LLM providers, self-hosted inference and MCP servers.
higress-group/higress
Cloud-native API gateway built on Istio and Envoy, with AI gateway features for LLM providers and MCP server hosting.
- Stars
- 9,619
- Forks
- 1,359
- Open issues
- 1,404
- Last commit
- 8 hours ago
What is it?
Higress is a cloud-native API gateway based on Istio and Envoy. It can be extended with Wasm plugins written in Go, Rust, or JS, and ships with a set of general-purpose plugins and a built-in management console.
On the AI side, Higress connects to mainstream domestic and international model providers through a unified protocol, and offers AI observability, multi-model load balancing, token rate limiting, and caching. It can also host MCP (Model Context Protocol) servers via its plugin mechanism, and includes an openapi-to-mcp tool for converting OpenAPI specifications into remote MCP servers.
Higress can serve as a Kubernetes ingress controller that supports the Ingress API and Gateway API, a microservice gateway that discovers services from registries such as Nacos, ZooKeeper, Consul, and Eureka, or a security gateway with WAF and authentication plugins. It can be run with a single Docker command for local use or deployed with Helm on Kubernetes.
Key features
- 01Wasm plugin extensions in Go, Rust, or JS with sandbox isolation and independent plugin upgrades
- 02Unified access to mainstream LLM providers with multi-model load balancing and token rate limiting
- 03Hosting of MCP servers, with openapi-to-mcp conversion for OpenAPI specs
- 04Support for both Kubernetes Ingress API and Gateway API
- 05Service discovery from Nacos, ZooKeeper, Consul, Eureka and integration with Dubbo and Sentinel
- 06Built-in WAF and authentication plugins such as key-auth, jwt-auth, and oidc
Use cases
- Putting a unified gateway in front of multiple LLM providers with observability and caching
- Hosting MCP servers so AI agents can call internal tools and APIs
- Replacing ingress-nginx as a Kubernetes ingress controller
- Acting as a microservice gateway for services registered in Nacos or Consul
Quick start
- 1
Create a working directory
mkdir higress; cd higress - 2
Start Higress with Docker (UI on 8001, gateway on 8080/8443)
docker run -d --rm --name higress-ai -v ${PWD}:/data -p 8001:8001 -p 8080:8080 -p 8443:8443 higress-registry.cn-hangzhou.cr.aliyuncs.com/higress/all-in-one:latest
Things to know
- Kubernetes and other installs require following the external Quick Start documentation.
- The default image registry is in China, so pulls may time out outside the region.
Topics
FAQ
- What is higress?
- Cloud-native API gateway built on Istio and Envoy, with AI gateway features for LLM providers and MCP server hosting. Higress is a cloud-native API gateway based on Istio and Envoy. It can be extended with Wasm plugins written in Go, Rust, or JS, and ships with a set of general-purpose plugins and a built-in management console.
- Is higress free to use? What license does it have?
- higress is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
- Is higress still maintained?
- Actively maintained. The last commit to higress was on Oct 22, 2026.
- What can I use higress for?
- Putting a unified gateway in front of multiple LLM providers with observability and caching Hosting MCP servers so AI agents can call internal tools and APIs Replacing ingress-nginx as a Kubernetes ingress controller Acting as a microservice gateway for services registered in Nacos or Consul
- How do I install higress?
- Create a working directory: mkdir higress; cd higress Start Higress with Docker (UI on 8001, gateway on 8080/8443): docker run -d --rm --name higress-ai -v ${PWD}:/data -p 8001:8001 -p 8080:8080 -p 8443:8443 higress-registry.cn-hangzhou.cr.aliyuncs.com/higress/all-in-one:latest
- Where can I get higress?
- The source code and installation instructions are on GitHub: https://github.com/higress-group/higress
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A Kubernetes-native API gateway implementing the Gateway API on top of Envoy, with support for AI gateway use cases.
An open source proxy for AI agents that adds security, observability and governance to LLM, MCP tool and agent-to-agent traffic.
Kong is a cloud-native, platform-agnostic API, LLM and MCP gateway, extensible through plugins, for routing and securing service traffic.
A high-performance, stateless, declarative API Gateway written in Go for routing, aggregating and securing calls to backend microservices.
An open source AI gateway and registry that federates MCP, A2A, and REST/gRPC APIs behind one endpoint with governance, discovery, and observability.