""" Cost calculator for DeepSeek Chat models. Handles prompt caching scenario. """ from typing import Tuple from litellm.litellm_core_utils.llm_cost_calc.utils import generic_cost_per_token from litellm.types.utils import Usage def cost_per_token(model: str, usage: Usage) -> Tuple[float, float]: """ Calculates the cost per token for a given model, prompt tokens, and completion tokens. Follows the same logic as Anthropic's cost per token calculation. """ return generic_cost_per_token( model=model, usage=usage, custom_llm_provider="deepseek" )