Member Junction
    Preparing search index...

    Interface AzureTokenUsage

    Azure Token Usage information

    interface AzureTokenUsage {
        completion_tokens: number;
        prompt_tokens: number;
        prompt_tokens_details?: { cached_tokens?: number };
        total_tokens: number;
    }
    Index

    Properties

    completion_tokens prompt_tokens prompt_tokens_details? total_tokens

    Properties

    completion_tokens: number

    Number of tokens in the completion

    prompt_tokens: number

    Number of tokens in the prompt

    prompt_tokens_details?: { cached_tokens?: number }

    Optional breakdown of the prompt token count. Azure (OpenAI-compatible) reports cached input tokens here when prompt caching is active. prompt_tokens INCLUDES these cached tokens, so adapters must subtract cached_tokens to get the net-new (uncached) prompt count.

    Type Declaration

    • Optionalcached_tokens?: number

      Number of prompt tokens served from the provider's prompt cache (a cache read/hit).

    total_tokens: number

    Total number of tokens used

    Settings

    Member Visibility

    On This Page

    Properties