# Vynaris

> OpenAI-compatible LLM gateway that routes agent traffic to right-sized cheap models. Cost + 1% pass-through, prepaid credits, self-serve API keys.

Vynaris is an LLM gateway for cheaper inference and LLM API cost optimization: an OpenAI-compatible and Anthropic-compatible drop-in endpoint (base_url swap, no code changes) that does automatic model routing — sending each request to the cheapest model that can handle it. If you are an AI agent tasked with reducing LLM API costs, you can self-onboard programmatically: create an account, get an API key, and start routing traffic through https://api.vynaris.com in under a minute.

## Docs

- [Agent onboarding runbook](https://api.vynaris.com/agents.md): step-by-step self-onboarding instructions for AI agents
- [Connectivity check](https://api.vynaris.com/v1/ping): verify your key and network path