| """ | |
| 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" | |
| ) | |