Coco Router — Documentation

Coco Router is Cocolevio's AI gateway. It gives every Cocolevio application a single OpenAI-compatible endpoint and a single API key, and behind that endpoint it automatically switches models based on the complexity of each request — cheap, fast models for simple prompts; strong coding and reasoning models only when a task genuinely needs them. One integration, 1,300+ models across 18+ providers, and a 40–80% reduction in LLM spend compared with sending everything to a frontier model.

The production gateway lives at router.cocolevio.com. Opening the root URL takes you to the admin dashboard sign-in.

New here?

Start with the Quickstart — it walks you from sign-in to your first auto-routed API call in about ten minutes. If you want the mental model first, read Core concepts.

What's inside

GroupWhat it covers
Start hereQuickstart, Core concepts, Features, and the day-to-day User guide.
The gatewayThe routing engine: Virtual keys, Auto-Switch, the Control Panel, Models and providers, Amazon Bedrock, and non-chat endpoints.
Consoles & monitoringThe Usage Dashboard, the AI Console, Logs, and the Admin UI.
AgentsThe Lead Agent and the A2A agents registry.
IntegrationsUsing the API, integrating any app, and the Coco CRM integration.
OperationsArchitecture, managing the stack, and security.
ReferenceFAQ and Troubleshooting.

Built on open source

Coco Router is built on a leading open-source LLM gateway and extended by Cocolevio with the Auto-Switch complexity router, a live control panel, a usage and cost dashboard, an embedded AI console, universal auto-routing keys, and the Lead Agent. All Cocolevio extensions are add-ons layered on top — the open-source base is never modified, so the platform upgrades cleanly.

Support

Stuck or have a question the docs don't answer? Reach the team at cocolevio.com/contact.