Supports text, vision, thinking, tool-use, function-calling, structured-outputs
Pricing is based on the number of tokens used, or other metrics based on the model type.