Higress

Active

Tagline:Open-source AI gateway for Agent development and LLM API management

Updated:September 6, 2026

Domain Rating:

Social Media:-

AI API Interface AI Developer Tools AI Agent Open Source API Available Self-Hosted For Developers For Enterprise Cloud-based Multi-language AI Agent Ready

Open Website

Visit Website

What is Higress?

Higress is a cloud-native API and AI gateway based on Istio and Envoy, extensible with Wasm plugins written in Go, Rust, and JS. It unifies governance of model APIs, MCP tool calls, and Kubernetes inference workloads, supporting 100+ mainstream LLM models with unified protocol conversion, load balancing, and fallback. A CNCF Sandbox project originated at Alibaba.

How to use Higress?

Run `curl -sS https://higress.cn/ai-gateway/install.sh | bash` to deploy the AI gateway locally via Docker. Access the console at http://localhost:8001/ to configure model providers, API keys, AI routes, and policies. For Kubernetes deployment, use Helm: `helm install higress -n higress-system higress.io/higress --create-namespace`.

Higress Core Features

Multi-Model Unified Access

Provides a unified gateway entry for OpenAI-compatible APIs and mainstream LLM services, shielding vendor authentication and interface differences. Supports model load balancing and fallback for reliability.

MCP Unified Management

Supports HTTP-to-MCP protocol conversion and native MCP service proxying, consolidating scattered tool capabilities into a unified entry point to reduce integration and maintenance costs.

Token Traffic Management

Beyond traditional QPS rate limiting, offers Token consumption quota management and rate limiting, combined with API Key pool rotation and consumer authentication to prevent model overload.

Semantic Caching

Supports precise and semantic caching of inference results, saving tokens and reducing latency for similar or repeated queries.

Cloud-Native API Gateway

Combines traffic, microservice, and security gateways in one. Supports mTLS mutual authentication, WAF integration, custom security plugins, and Nginx Ingress annotation compatibility for low-cost migration.

Plugin Extensibility

Quickly integrates RAG, vector databases, logging, observability, and caching through a hot-pluggable plugin ecosystem. Supports custom Wasm plugin development in Go, Rust, and JS.

Higress Use Cases

Multi-model proxying for unified LLM API access across vendors

Application protection with data privacy and content filtering for AI interactions

MCP Server hosting for AI Agents to call external tools and services

Token management and quota control for AI application consumers

Higress Frequently Asked Questions

Is Higress ready for production systems?

Yes, it is recommended to use v1.0.0 and above. Refer to the project roadmap for version planning and feature stability.

What is the difference between Higress and Envoy Gateway?

Higress is built on Envoy and follows the Gateway API standard like Envoy Gateway, but additionally provides WAF protection, authentication plugins, multi-registry support, protocol conversion (HTTP-to-Dubbo), and WASM/Lua custom plugin support.

Can Higress run without Kubernetes, using only Docker?

Yes, Higress can be started with a single Docker command for local development and testing, with no Kubernetes required.

What is the relationship between Higress and Alibaba Cloud MSE Gateway?

The MSE Cloud Native Gateway is the commercial version of Higress, with additional enterprise features and managed service capabilities.

Does Higress support elastic scaling? Is the gateway stateless?

Yes, Higress supports elastic scaling and the gateway is stateless, making it suitable for horizontal scaling in production environments.

Higress Product Background & Official Links

Higress Support Email & Customer service contact & Refund contact etc.

More Contact, visit the about us page(https://higress.ai/docs/latest/overview/what-is-higress)

Higress Company Info

Higress Company name: LF Projects, LLC

Higress Developer Type: company

More about Higress, Please visit the about us page(https://higress.ai/docs/latest/overview/what-is-higress)

Higress Pricing

Higress Pricing Link: https://higress.ai/ai-gateway

Higress GitHub

Higress GitHub Link: https://github.com/alibaba/higress

Higress Pricing & Plans

Community Edition

Free

Complete AI gateway core capabilities
Community support and documentation
Self-hosted with Docker or Kubernetes

Subscribe & upgrade now

Serverless

Recurring

Starts at ¥0, pay-as-you-go

Fully managed, no ops required
Auto-scaling based on call volume
Unified LLM API, Agent API, and MCP Server support
99.95% SLA with Enterprise edition

Subscribe & upgrade now

Enterprise

contact

Contact sales

Performance and resources matched to business scale
Custom SLA negotiation
Dedicated entry, custom domain, and transport encryption

Subscribe & upgrade now

Higress Tags

Open Source API Available Self-Hosted For Developers For Enterprise Cloud-based Multi-language AI Agent Ready

Best Suited Occupations for Higress

Software Developers IT Ops & Security Product Managers

What is Higress?

How to use Higress?

Higress Core Features

Multi-Model Unified Access

MCP Unified Management

Supports HTTP-to-MCP protocol conversion and native MCP service proxying, consolidating scattered tool capabilities into a unified entry point to reduce integration and maintenance costs.

Token Traffic Management

Beyond traditional QPS rate limiting, offers Token consumption quota management and rate limiting, combined with API Key pool rotation and consumer authentication to prevent model overload.

Semantic Caching

Supports precise and semantic caching of inference results, saving tokens and reducing latency for similar or repeated queries.

Cloud-Native API Gateway

Plugin Extensibility

Quickly integrates RAG, vector databases, logging, observability, and caching through a hot-pluggable plugin ecosystem. Supports custom Wasm plugin development in Go, Rust, and JS.

Higress Frequently Asked Questions

Is Higress ready for production systems?

Yes, it is recommended to use v1.0.0 and above. Refer to the project roadmap for version planning and feature stability.

What is the difference between Higress and Envoy Gateway?

Can Higress run without Kubernetes, using only Docker?

Yes, Higress can be started with a single Docker command for local development and testing, with no Kubernetes required.

What is the relationship between Higress and Alibaba Cloud MSE Gateway?

The MSE Cloud Native Gateway is the commercial version of Higress, with additional enterprise features and managed service capabilities.

Does Higress support elastic scaling? Is the gateway stateless?

Yes, Higress supports elastic scaling and the gateway is stateless, making it suitable for horizontal scaling in production environments.