Interface TokenUsage

Token counts reported by an LLM provider.

interface TokenUsage {
    cachedInputTokens?: number;
    inputTokens?: number;
    outputTokens?: number;
    reasoningTokens?: number;
    totalTokens?: number;
}

Index

Properties

cachedInputTokens? inputTokens? outputTokens? reasoningTokens? totalTokens?

Properties

`Optional`cachedInputTokens

cachedInputTokens?: number

Input tokens served from a provider cache.

`Optional`inputTokens

inputTokens?: number

Tokens read from the request.

`Optional`outputTokens

outputTokens?: number

Tokens generated in the response.

`Optional`reasoningTokens

reasoningTokens?: number

Tokens used for model reasoning.

`Optional`totalTokens

totalTokens?: number

Total input and output tokens.