Hadrian Gateway
A high-performance, production-ready AI Gateway for LLMs
Hadrian is an AI Gateway that provides a unified OpenAI-compatible API for routing requests to multiple LLM providers.
Key Features
- OpenAI-Compatible API - Drop-in replacement for OpenAI's API, including OpenResponses
- Multi-Provider Support - Route to OpenAI, Anthropic, Bedrock, Vertex, Azure, and more
- Built-in Web UI - Chat interface with multi-model support
- Multi-Tenancy - Organizations, teams, projects, and users
- Budget Enforcement - Cost tracking and limits per project/user
- API Key & OIDC Auth - Flexible authentication options
- RAG / Vector Stores - File search with automatic document processing
- Agents & Shell Tool - Server-side shell and MCP via the Responses API
- Observability - Logging, metrics, and tracing
Quick Start
Method
OS
Libc
Features
Production deployment
Model catalog
Embedded docs
CSV export
JSON schema
SAML SSO
Kreuzberg OCR
ClamAV scanning
Microsandbox containers
OpenSandbox containers
curl -L \
https://github.com/ScriptSmith/hadrian/releases/latest/download/hadrian-x86_64-unknown-linux-musl-standard.tar.gz \
| tar xz
./hadrianDocumentation Sections
| Section | Description |
|---|---|
| Getting Started | Installation and basic setup |
| Configuration | All configuration options |
| Features | Detailed feature guides |
| API Reference | OpenAI-compatible API documentation |
| Deployment | Production deployment guides, build profiles (full, headless, standard, ...) |
| Security | Security best practices |
| Troubleshooting | Common issues and solutions |