AgentOptimizationJobTokenUsage interface

Aggregated token usage for one optimization stage and model.

Properties

cached_tokens	Input tokens served from the model's cache. Included in input_tokens and not additive. If omitted, the service defaults to 0.
input_tokens	Total input tokens. Omitted when unmeasured.
model	Model name or deployment name that served the calls.
output_tokens	Total output tokens. Omitted when unmeasured.
reasoning_tokens	Reasoning tokens counted separately by reasoning models. Included in output_tokens and not additive. If omitted, the service defaults to 0.
stage	Optimization stage that generated these calls.
total_tokens	Sum of input_tokens + output_tokens. Omitted when unmeasured.

Input tokens served from the model's cache. Included in input_tokens and not additive. If omitted, the service defaults to 0.

cached_tokens?: number

number

Total input tokens. Omitted when unmeasured.

input_tokens?: number

number

Model name or deployment name that served the calls.

model?: string

string

Total output tokens. Omitted when unmeasured.

output_tokens?: number

number

Reasoning tokens counted separately by reasoning models. Included in output_tokens and not additive. If omitted, the service defaults to 0.

reasoning_tokens?: number

number

Optimization stage that generated these calls.

stage: AgentOptimizationStage

Sum of input_tokens + output_tokens. Omitted when unmeasured.

total_tokens?: number

number