sglang-model-gateway

Official

Rust gateway in front of vLLM/SGLang inference.

Authorair-gapped
Version1.0.0
Installs0

System Documentation

What problem does it solve?

The sglang-model-gateway provides a Rust-based front door for OpenAI-compatible HTTP and gRPC workloads, routing requests to vLLM and SGLang inference workers behind Kubernetes deployments with predictable latency and routing behavior.

Core Features & Use Cases

  • Deterministic, policy-driven routing (cache_aware, random, round_robin, power_of_two, prefix_hash, manual)
  • Kubernetes-ready deployment patterns: per-model deployments, mesh-enabled HA, and service-discovery or static worker URLs
  • Air-gapped/offline operation with local model mirrors and tokenizer/model files

Quick Start

Run sgl-model-gateway in front of your workers and configure a simple setup with a static worker list and cache_aware policy to preserve prefix locality.

Dependency Matrix

Required Modules

None required

Components

assetsreferences

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: sglang-model-gateway
Download link: https://github.com/air-gapped/skills/archive/main.zip#sglang-model-gateway

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 510,000+ vetted skills library on demand.