API Reference
OpenAI-compatible API documentation with interactive examples
Hadrian Gateway provides an OpenAI-compatible API with extensions for multi-provider support and multi-tenancy.
Base URL
http://localhost:8080Authentication
The gateway supports API key and JWT authentication. See the Authentication page for details on authenticating API requests.
Public API (OpenAI-Compatible)
These endpoints follow the OpenAI API specification and work with existing OpenAI client libraries.
Chat Completions
Create chat completions with streaming, tool use, and vision support
Completions
Legacy text completions API
Embeddings
Generate vector embeddings for semantic search
Models
List available models from all providers
Images
Generate images with DALL-E
Audio
Text-to-speech and transcription
Files
Upload files for vector stores
Vector Stores
RAG with semantic search
Admin API
RESTful endpoints for managing multi-tenant configuration. All admin endpoints require authentication.
Organizations
Top-level multi-tenancy container
Projects
Workload separation within organizations
Teams
Group users for permission management
Users
User management and membership
API Keys
Create and manage API keys
Dynamic Providers
Runtime LLM provider configuration
Usage
Query usage statistics and costs
Model Pricing
Configure per-model pricing
Conversations
Chat history storage
Prompts
Reusable prompt templates
Audit Logs
Security audit trail
Access Reviews
Compliance reporting
Me
Self-service user endpoints
Interactive Documentation
The gateway also provides interactive API documentation at runtime:
- Scalar UI:
/api/docs- Interactive API explorer - OpenAPI Spec:
/api/openapi.json- Machine-readable specification