One gateway.
Any AI model.
Total control.
The only LLM gateway that uses AI to decide which AI to use.
Drop Routerly between your app and your LLM providers. Intelligent routing, cost tracking, budget enforcement — zero code changes.
Open source · Built with TypeScript · Runs on Docker · AGPL-3.0 licensed · Supports OpenAI, Anthropic, Gemini, Mistral, Ollama, Cohere, xAI
Install
Get started in seconds
Pick your install method and you're ready to go.
# Works everywhere. Installs everything. You're welcome. 🔥
Works on macOS, Windows & Linux. The one-liner installs Node.js and everything else for you.
Watch
Meet Routerly
See how Routerly routes your AI requests to the best model — automatically.
Any client
Any model
How it works
The missing layer between your code
and every AI provider
Today, your app talks directly to a single AI provider — OpenAI, Anthropic, or whichever you chose first. That means you're locked in: if their API goes down, you go down. If their prices change, your bill changes. If you want to try a different model, you rewrite code.
Routerly sits in the middle. You point your app — or any AI tool like Cursor, Open WebUI, or LangChain — at localhost:3000 instead of the provider's URL. From that moment, Routerly takes over: it picks the best available model for each request, enforces your budget, tracks every token spent, and automatically reroutes if a provider fails.
Your code doesn't change. Your team doesn't need to know. You just get full control.
Features
Everything you need
to run AI at scale
Built for developers who need reliability, cost visibility, and full control over their AI infrastructure.
Intelligent Routing
9 configurable policies score every candidate in parallel. Pick cheapest, fastest, healthiest, most capable — or combine them.
Drop-in Compatible
Implements OpenAI and Anthropic wire formats with full streaming. Swap a base URL — your client code is unchanged.
Cost Tracking
Per-token pricing on every call. Built-in presets for GPT-4o, Claude, Gemini. Custom pricing for any model.
Budget Enforcement
Hourly, daily, weekly, or monthly caps per project or globally. Over-budget models are filtered before routing starts.
Project Isolation
Separate Bearer tokens per project, each with its own model list, routing config, and budget limits.
Web Dashboard
Visual analytics, model management, project config, and user & role admin — no CLI needed.
Use Cases
Built for every AI workflow
From indie projects to enterprise deployments — Routerly adapts to how you work.
SaaS & multi-tenant
One project per tenant, hard spend caps, automatic cheapest-model routing. No code changes, no billing surprises.
Local-first development
Develop against Ollama locally, promote to GPT-4o in production. Same API, same code — swap one token.
Automatic cost reduction
Route simple tasks to cheap models, complex ones to capable models. Set hard limits hourly, daily, or per request.
Resilience & automatic failover
Register the same capability across OpenAI, Claude, and Gemini. Routerly detects failures and reroutes in real time.
AI tools & chat UIs
Cursor, Open WebUI, OpenClaw, LibreChat — any tool that speaks OpenAI or Anthropic format. One endpoint to rule them all.
Model evaluation & A/B testing
Split traffic with the fairness policy. Compare cost, latency, and error rate side by side in the dashboard.
Providers
All major providers. One endpoint.
OpenAI, Anthropic, Gemini, Ollama, Mistral, Cohere, xAI — or any OpenAI-compatible custom endpoint.
| Provider | API Key | Streaming | Pricing Presets |
|---|---|---|---|
| OpenAI | ✓ | ✓ | ✓ |
| Anthropic | ✓ | ✓ | ✓ |
| Google Gemini | ✓ | ✓ | ✓ |
| Ollama (local) | — | ✓ | — |
| Mistral | ✓ | ✓ | ✓ |
| Cohere | ✓ | ✓ | ✓ |
| xAI (Grok) | ✓ | ✓ | ✓ |
| Custom HTTP | — | ✓ | — |
Comparison
How Routerly compares
Routerly is the only self-hosted gateway with LLM-powered routing and native support for both OpenAI and Anthropic formats — with zero infrastructure dependencies.
| Feature | Routerly | LiteLLM | OpenRouter |
|---|---|---|---|
| Self-hosted | ✓ | ✓ | — |
| Data stays on your infra | ✓ | ✓ | — |
| Native Anthropic API format | ✓ | — | — |
| LLM-powered smart routing | ✓ | — | — |
| No database required | ✓ | — | N/A |
| Local models (Ollama) | ✓ | ✓ | — |
| Bring your own API keys | ✓ | ✓ | — |
| Built-in web dashboard | ✓ | ✓ | ✓ |
Routerly is the only option where the gateway itself is intelligent — LLM-powered routing adapts to every request in real time.