Skip to content
alfe.ai

Models & routing overview

Alfe gives every agent access to many large language model providers through a single endpoint. You don’t sign up with each provider, juggle separate API keys, or reconcile separate bills — your agent picks a model, and Alfe routes the request to the right provider on your behalf. All usage draws from your account’s shared credit pool.

One account, many providers

Section titled “One account, many providers”

Every Alfe account can use models from all of the supported providers. There’s nothing extra to enable per provider — the full catalog is available as soon as you have an agent, and you choose which model to use per agent (and even per request).

The supported model families are:

Provider Models
Anthropic Claude family
OpenAI GPT family
Google Gemini family
DeepSeek DeepSeek family
MiniMax MiniMax family
Mistral Mistral and Codestral
xAI Grok family
Zhipu AI GLM family
OpenRouter 300+ additional models from many providers

The exact list of models — and which are enabled for your account — is shown in the dashboard when you configure an agent, and it can change over time as new models are added. New models can appear without any action on your part.

How routing works

Section titled “How routing works”

You reference a model by its name (for example, a Claude, GPT, or Gemini model). Alfe forwards the request to the provider that serves that model and streams the response back. Because everything goes through one endpoint:

  • You can switch models without touching your integrations. Changing the model an agent uses doesn’t require reinstalling or reconfiguring anything else.
  • You can mix models across requests. The same agent can use a fast, cheap model for routine work and a stronger model for hard problems.
  • Usage is metered centrally and billed from your credit pool, so there’s a single place to see what you’re spending.

What’s next

Section titled “What’s next”