Skip to content

Cloudflare AI Gateway lets you call a pool of models by one name

Model aliases route a single name like custom/support-fast to several underlying models, with fallback ordering or smart routing handled on the gateway.

Source: Cloudflare

Cloudflare’s AI Gateway added model aliases on 4 November. An alias is a pool of models your application calls by one name, such as custom/support-fast, instead of a specific model ID.

Aliases are created on a gateway’s Routing page and work one of two ways: fallback ordering, where the gateway tries models in sequence and moves on if one fails, or smart routing, where the gateway picks which model handles each request. Calling one is the same as any other model request, just with the alias as the model name:

{
  "model": "custom/support-fast",
  "input": {
    "messages": [{ "role": "user", "content": "Reply with a short greeting." }]
  }
}

Aliases are scoped to the gateway specified in the cf-aig-gateway-id header.

Why it matters: this decouples your application code from the specific model behind it. Swap a model, reorder fallbacks, or change the routing logic entirely, all from the gateway, with nothing to redeploy on your end.

Share

Related reading

ML for Web Devs

Tokens, context windows, and why your bill is what it is

The unit of billing for language models is not the request or the word. Once you understand tokens and how a context window fills up, the invoice stops being a surprise.

ML for Web Devs

RAG is just search with extra steps

Retrieval-augmented generation gets written up like an architecture. It is a search query, a string concatenation and one API call, and the search query is the part that decides whether it works.

ML for Web Devs

A prompt is a spec, so write it like one

Prompt engineering has very little to do with magic words. It is the same skill as writing a clear ticket for a capable contractor who will not ask you any questions.

More from The Wire

All briefs