Skip to content
Erebrosdocs

PROVIDERS

PROVIDERS is a single Chat engine over the model aggregators, in Chat, /chat?engine=providers.

PROVIDERS, Erebros

What it does

Section titled “What it does”

Wraps three aggregators, FreeInference, OpenRouter and TokenRouter, behind one engine, exposing hundreds of models including free tiers. Any single aggregator key turns the engine on. It also joins the fleet as a global-model family: a pick is addressed as providers::<aggregator>::<model>, so an aggregator model can be set as the global model in the top bar.

Key actions

Section titled “Key actions”
  • Pick an aggregator and model, from the combined catalog of hundreds of models.
  • Use free tiers, many models are available at no cost.
  • Set as global model, any aggregator model can drive the default-engine apps (Loop, Pipeline).

Setup

Section titled “Setup”

Set at least one aggregator key in .env.local, then restart the server:

  • FREEINFERENCE_API_KEY (FreeInference also serves /v1/models keyless)
  • OPENROUTER_API_KEY
  • TOKENROUTER_API_KEY (plus TOKENROUTER_URL)

Any one key is enough to enable the engine; add more to widen the catalog.

Tips

Section titled “Tips”
  • FreeInference exposes its model list without a key, so the catalog can be browsed before you commit a key.

See also

Section titled “See also”