RainyModel: LLM Routing Proxy

Part of the Orcest AI Ecosystem

---

RainyModel is an intelligent LLM routing proxy that provides OpenAI-compatible API endpoints with automatic routing across free, internal, and premium providers. It is the central LLM orchestration layer of the Orcest AI ecosystem.

Orcest AI Ecosystem

Service	Domain	Role
Lamino	llm.orcest.ai	LLM Workspace
RainyModel	rm.orcest.ai	LLM Routing Proxy
Maestrist	agent.orcest.ai	AI Agent Platform
Orcide	ide.orcest.ai	Cloud IDE
Login	login.orcest.ai	SSO Authentication

Features

• Smart Routing: Automatic routing chain: FREE (HF) -> INTERNAL (Ollama) -> PREMIUM (OpenRouter)
• OpenAI-Compatible API: Drop-in replacement for any OpenAI-compatible client
• Model Aliases: rainymodel/auto, rainymodel/chat, rainymodel/code, rainymodel/agent
• Dual Ollama Backends: 16GB primary (qwen2.5:14b) + 8GB secondary (qwen2.5:7b)
• Policy Headers: X-RainyModel-Policy for routing control (default/uncensored/premium)
• Observability: Response headers with route, upstream, model, latency info
• Circuit Breaker: Automatic failover on upstream errors
• Rate Limiting: Per-user API key rate limits

API Endpoints

POST /v1/chat/completions  - Chat completions (OpenAI-compatible)
GET  /v1/models            - List available model aliases
GET  /health               - Health check

Model Aliases

Alias	Use Case	Routing Priority
rainymodel/auto	General purpose	HF -> Ollama -> OpenRouter
rainymodel/chat	Conversational	HF -> Ollama -> OpenRouter
rainymodel/code	Code generation	HF Coder -> Ollama Coder -> OpenRouter
rainymodel/agent	Agent tasks	HF -> Ollama -> OpenRouter

Response Headers

Header	Description
x-rainymodel-route	free / internal / premium
x-rainymodel-upstream	hf / ollama / openrouter
x-rainymodel-model	Actual model used
x-rainymodel-latency-ms	Request latency

Deployment

Deployed on Render with auto-deploy from main branch. See render.yaml for configuration.

License

This project is licensed under the MIT License.

Part of the Orcest AI ecosystem.