# Deploy OmniRoute with Easypanel

URL: /templates/omniroute

OmniRoute is an MIT-licensed, OpenAI-compatible AI gateway that gives you one endpoint for hundreds of providers and models. It auto-routes and falls back across your subscriptions and API keys (Subscription → API Key → Cheap → Free tiers), so coding tools like Claude Code, Cursor, Codex, and Cline keep working even when a provider hits a rate limit or quota. It ships a dashboard for managing providers, keys, and subscription OAuth logins, with live usage analytics, token-compression, and circuit breakers to avoid dead keys. It is not a LiteLLM replacement - LiteLLM is a YAML/model_list config layer, while OmniRoute is a dashboard plus subscription OAuth with automatic multi-tier fallback.

## Installation notes

Open the dashboard at your domain and log in with the admin password shown in this service's environment variables (INITIAL_PASSWORD) - change it immediately under Settings → Security. Then add a provider or connect a subscription via OAuth from the dashboard to start routing requests through http://your-domain/v1.

## Benefits

### One Endpoint, Hundreds of Providers

Point any OpenAI-compatible tool at a single OmniRoute URL and reach 350+ providers and 1200+ models without juggling separate configs.

### Never Lose a Session to a Rate Limit

Automatic 4-tier fallback (Subscription, API Key, Cheap, Free) keeps requests flowing when one provider throttles, errors, or runs dry.

### Self-Hosted, MIT-Licensed

Full control over your data, keys, and routing rules, with an open-source core instead of a hosted black box.

## Features

### Dashboard With Subscription OAuth

Connect provider subscriptions and manage API keys through a web dashboard, including device-flow OAuth logins for supported providers.

### Smart Routing & Circuit Breakers

19 routing strategies, key cooldowns, and circuit breakers route around dead keys and failing providers automatically.

### Token Compression

Built-in prompt compression can cut token usage by 15-95% on tool-heavy coding-agent traffic.

### Live Usage Analytics

Track spend, quota consumption, and p95 latency per provider and key from the dashboard.

### MCP & A2A Agent Protocols

Expose OmniRoute itself to agents over MCP or A2A so they can manage routing, providers, and combos autonomously.

## Configuration

- `appServiceName` (App Service Name, required): omniroute

- `appServiceImage` (App Service Image, required): diegosouzapw/omniroute:3.8.51

## Services

- omniroute: app (diegosouzapw/omniroute:3.8.51)

- omniroute-redis: redis

## Resources

- [Website](https://omniroute.online)

- [GitHub](https://github.com/diegosouzapw/OmniRoute)

- [Documentation](https://github.com/diegosouzapw/OmniRoute/blob/main/docs/README.md)

- [Template source](https://github.com/easypanel-io/templates/tree/main/templates/omniroute)