Skip to main content
Raven maintains a catalog of available models across all providers. The catalog includes pricing, capabilities, and context window information to help you choose the right model.

​
Model Catalog

Navigate to Models in the dashboard to browse all available models. You can filter by:
  • Provider — Show models from specific providers
  • Capability — Filter by chat, function calling, vision, etc.
  • Context window — Filter by minimum context length

​
Model Metadata

Each model in the catalog includes:
FieldDescription
ProviderThe LLM provider (e.g., OpenAI, Anthropic)
Model IDThe identifier used in API requests
Input PriceCost per 1M input tokens
Output PriceCost per 1M output tokens
Context WindowMaximum token capacity
CapabilitiesSupported features (chat, vision, tools)

​
Using Models

Specify the model in your API request: 
{
  "model": "gpt-4o",
  "messages": [{"role": "user", "content": "Hello"}]
}
Raven automatically routes the request to the provider that serves that model. If you have multiple providers configured for the same model, routing rules determine which one is used.

​
Model Syncing

Raven periodically syncs its model catalog from providers to stay up to date with the latest models and pricing. You can also trigger a manual sync from the admin panel.

​
Cost Tracking

Every request logs the model used along with token counts:
  • Input tokens — Tokens in the prompt
  • Output tokens — Tokens in the response
  • Reasoning tokens — Tokens used for extended thinking (e.g., Claude)
  • Cached tokens — Tokens served from provider-level cache
Costs are calculated using the model’s current pricing and displayed in the analytics dashboard.