@chinmaymk/aikit / GenerationUsage
Interface: GenerationUsage
Defined in: types.ts:123
Usage information for a generation request. Like a receipt from your AI shopping spree, but with tokens instead of groceries.
Properties
inputTokens?
optionalinputTokens:number
Defined in: types.ts:125
The number of tokens in the input/prompt. What you fed the AI to get it thinking.
outputTokens?
optionaloutputTokens:number
Defined in: types.ts:127
The number of tokens in the generated output. The AI's brilliant (or not so brilliant) response.
totalTokens?
optionaltotalTokens:number
Defined in: types.ts:129
The total number of tokens used in the request (input + output). Your grand total bill.
reasoningTokens?
optionalreasoningTokens:number
Defined in: types.ts:131
The number of tokens used for reasoning (o-series models). The AI showing its work, like a good student.
cacheTokens?
optionalcacheTokens:number
Defined in: types.ts:133
The number of tokens used for cached content (if applicable). Free tokens! The AI remembered something.
timeToFirstToken?
optionaltimeToFirstToken:number
Defined in: types.ts:135
Time to first token in milliseconds. How long you waited for the AI to start thinking out loud.
totalTime?
optionaltotalTime:number
Defined in: types.ts:137
Total time for the entire generation in milliseconds. How long you waited for the AI to finish its masterpiece.