Skip to main content
Welcome to Open Next Router! This guide will help you install, configure, and route your first LLM request through the gateway locally.
Prerequisites: Ensure you have Go 1.25+ and Make installed on your system. If you prefer Docker, check out the Docker Guide.

Getting Up and Running

1

Clone the Repository

First, grab the source code from GitHub:
2

Build the Binary

Compile the onr binary using the provided Makefile:
This will place the executable in the bin/ directory or your $GOPATH/bin. If it’s globally available, you can simply run onr.
3

Initialize Configuration Files

Open Next Router relies on three main YAML configuration files. You can start with the defaults provided in the repository.
Main server configurations including ports, logger, and paths to other config files.
Defines your downstream API keys, rate limits, and which DSL routing profile to apply.
Safely store the actual upstream API keys (e.g., OpenAI, Anthropic tokens) required to dispatch requests.
4

Run Configuration Tests

It’s always a good idea to validate your configuration syntax before starting the server:
If everything is correct, you’ll see a Configuration is valid message.
5

Start the Gateway

Launch the router:
You should see the server boot up and begin listening on http://localhost:3300.
6

Send Your First Request

Open Next Router natively exposes an OpenAI-compatible /v1/chat/completions endpoint. Try calling it:
Replace your-onr-client-token with a token defined in your keys.yaml.

Next Steps

Now that you have traffic flowing, explore these essential topics to unlock the power of ONR:

DSL Overview

Learn how to write dynamic routing rules, rewrites, and token tracking.

Load Balancing & Fallbacks

Configure multi-provider failover for maximum resilience.

Observability

Set up Promtail, Loki, and Grafana for comprehensive gateway monitoring.

Admin Panel

Manage keys and live DSL configurations visually.