Models
How Troviale's model catalog works — model IDs, routing across providers, and where to find pricing and status.
A model is identified by a stable ID, such as gpt-4o or claude-sonnet-4-6.
You select a model with the model field on each request, and the gateway
routes to the correct provider behind the scenes.
Selecting a model
The model field is the only thing that changes when you switch providers:
# OpenAI
client.chat.completions.create(model="gpt-4o", messages=...)
# Anthropic — same client, same code
client.chat.completions.create(model="claude-sonnet-4-6", messages=...)Because the request and response shapes are normalized to the OpenAI format, your code does not need to know which provider served the response.
Catalog, pricing, and status
The full list of available models — with per-token pricing, context windows, capabilities, and current status — lives in the model catalog. Each model also has a dedicated page with its details.
Status
Models carry a status:
- Active — generally available and recommended for production.
- Preview — available for testing; behavior or pricing may change.
- Deprecated — still callable, but scheduled for removal. Migrate away.
Check the catalog before relying on a model in production.