Default WorkspaceGuides
429 fallback, coding agents & gateway comparisons
API DocsBYOK & Keys

LLM Router: What It Is, Free and BYOK Options, and How to Choose (2026)

Updated · By Hisham · Facts and limits are dated; check the linked sources.

Short answer: an LLM router is a layer between your app and the model providers that decides, for each request, which model and provider should answer it, and tries another one when the first fails. You call one API; the router handles model choice, provider keys, fallbacks and limits. Most LLM routers expose an OpenAI-compatible endpoint, so switching to one is usually a base URL change.

This page is the hub for everything we have written about LLM routers: what they do, how free and bring-your-own-key (BYOK) routers differ, and links to every comparison and guide on this site. We build SentioGate, an AI LLM router, so we say where it fits and where another tool is the better pick.

As of 2026-10-31. Competitor facts on the linked pages are dated and sourced.

What an LLM router does

  • One API for many models. Your code sends OpenAI-style chat requests to one endpoint; the router translates and forwards them.
  • Model selection. Some routers let you name a model; some, like SentioGate's model: "auto", pick a capable model per request.
  • Fallback. When a provider returns 429 or 5xx, the router retries elsewhere. SentioGate automatically falls back to another model if one fails.
  • Key and spend management. Provider keys live in the router instead of every service, with usage tracking per key.
  • Observability. Request logs, latency and token usage, sometimes exported to tools such as Langfuse.

An LLM router overlaps with an LLM gateway; the what is an LLM gateway explainer covers the difference.

Free LLM routers

A free LLM router lets you send requests without paying for tokens, usually from a shared pool or from providers' own free tiers. Options include OpenRouter's :free model variants (20 requests per minute, with a daily cap; see the OpenRouter free models guide) and SentioGate's free community pool: send model: "auto" with no provider keys, no credit card, 600 requests per 5 hours per user, a shared community evaluation tier. Free routing is for evaluation and prototypes. See the free LLM API list for provider free tiers and the free chatbot API guide for building on one.

BYOK LLM routers

A BYOK LLM router uses your own provider API keys, so each provider bills you directly at its own prices. The router's business model then matters: SentioGate charges 0% token markup on BYOK traffic, with up to 5 keys per provider. OpenRouter charges 5.5% on credit purchases and also supports BYOK; Requesty, Portkey and LiteLLM support your own keys too. The BYOK LLM gateway guide explains the model in detail.

Hosted vs self-hosted LLM routers

A hosted router (OpenRouter, Requesty, Portkey's cloud, SentioGate) runs for you on someone else's servers; you get an endpoint and a dashboard. A self-hosted router such as LiteLLM runs on your own infrastructure, which gives you full control and a server to operate. SentioGate is hosted only, running on a global edge network. If you must keep the router inside your own network, a self-hosted option is the better pick; see LiteLLM vs Bifrost vs Helicone vs SentioGate and the LiteLLM alternative guide.

How to choose an LLM router

  1. Who pays the provider? Router credits (marketplace) or your own keys (BYOK).
  2. What does it cost on top? A fee on credits, a per-request or per-log plan, or no markup.
  3. Where must it run? Hosted, or self-hosted on your own infrastructure.
  4. What happens on failure? Automatic fallback across providers, or you handle retries.
  5. What do you need beyond routing? Governance, an MCP gateway, prompt management or deep observability point to Portkey or Helicone. SentioGate has no MCP gateway.

Every LLM router comparison

Every LLM router guide

Choosing a router or gateway

Free tiers

Rate limits and failover

Cost and integrations

FAQ

What is an LLM router?

An LLM router sits between your app and model providers. You call one API, and the router picks the model and provider for each request, handles provider keys and tries another provider when one fails.

Is there a free LLM router?

Yes, for evaluation. OpenRouter has free model variants with per-minute and daily limits, and SentioGate's free community pool gives 600 requests per 5 hours per user with no credit card. Free routing is not meant for production traffic.

What is a BYOK LLM router?

A BYOK (bring your own key) router sends requests with your own provider API keys, so each provider bills you directly. SentioGate charges 0% token markup on BYOK traffic; other routers, including OpenRouter, Requesty and LiteLLM, also support your own keys.

Should I use a hosted or a self-hosted LLM router?

Use a hosted router if you want an endpoint without running servers. Use a self-hosted router such as LiteLLM if it must run inside your own infrastructure. SentioGate is hosted only.

Sources

All accessed 31 Oct 2026.