Skip to main content
The model catalog shows the models available to your account, including provider, capability labels, context length, and pricing. Use the model name exactly as shown in the catalog when setting the model field.

Model name

Copy the exact model identifier from the catalog. Names are case-sensitive.

Token pricing

Most models charge separately for input and output tokens.

Usage logs

Review request status, tokens, and cost after each call.

​
Pricing modes

​
Token-based models

cost = (input tokens x input price + output tokens x output price) x group multiplier
Prices are usually listed per million tokens. Some multimodal or task-based models charge per call instead of by token count.

​
Cost controls

  • Choose smaller models for simple tasks.
  • Limit context to relevant messages and documents.
  • Set max_tokens when output should be bounded.
  • Use separate API keys with quota limits for each project.
  • Review usage logs regularly.

​
Next steps