Skip to main content

Interface: TokenUsage

Defined in: packages/agentos/src/api/generateText.ts:91

Token consumption figures reported by the provider for a single completion call. All values are approximate and provider-dependent.

Properties​

cacheCreationTokens?​

optional cacheCreationTokens: number

Defined in: packages/agentos/src/api/generateText.ts:120

Tokens written to the provider's prompt-prefix cache as a new cache entry. Billed at the cache-creation rate (1.25× input price on Anthropic for 5-minute TTL, 2× for 1-hour TTL). NOT also counted in promptTokens. A cacheReadTokens of 0 and cacheCreationTokens > 0 indicates the first call that filled the cache; subsequent calls with a cache hit flip the numbers.

cacheReadTokens?​

optional cacheReadTokens: number

Defined in: packages/agentos/src/api/generateText.ts:111

Tokens served from the provider's prompt-prefix cache. When present, these were billed at the cache-read rate (0.1× input price on Anthropic) and are NOT also counted in promptTokens. Callers that want total tokens-ever-sent should add `promptTokens + cacheReadTokens

  • cacheCreationTokens`.

Undefined when the provider does not report cache usage (OpenAI's auto-cache does not expose this at the per-call layer; Anthropic does via cache_read_input_tokens).

completionTokens​

completionTokens: number

Defined in: packages/agentos/src/api/generateText.ts:95

Number of tokens in the model's response.

costUSD?​

optional costUSD: number

Defined in: packages/agentos/src/api/generateText.ts:99

Total cost reported by the provider across all steps, when available.

promptTokens​

promptTokens: number

Defined in: packages/agentos/src/api/generateText.ts:93

Number of tokens in the prompt / input sent to the model.

totalTokens​

totalTokens: number

Defined in: packages/agentos/src/api/generateText.ts:97

Sum of promptTokens and completionTokens.