The Model Registry is Infere's live catalog of every available model, with capabilities, context window, max output, and per-1M-token pricing. A nightly sync pulls data from provider APIs, so the registry - not a marketing page - is the source of truth for what is available and what it costs.
Why it matters
Model names, capabilities, and prices change constantly. A live catalog prevents both stale marketing claims and stale application assumptions.
How it works
A nightly job pulls model data from provider APIs and enriches it with pricing and metadata, publishing the catalog to the edge. The registry UI reads the same data the gateway uses to route and price requests.