OperationTokenUsageDto
Tokens an AI operation consumed, as recorded in the operation metadata. A kind the provider did not report is 0.
| Name | Type | Description | Notes |
|---|---|---|---|
| totalTokens | Long (int64) | All tokens of the request: prompt plus completion. | [optional] [example: 20747] |
| promptTokens | Long (int64) | Tokens sent to the model, cached ones included. | [optional] [example: 19332] |
| completionTokens | Long (int64) | Tokens the model generated, reasoning ones included. | [optional] [example: 1415] |
| cachedTokens | Long (int64) | Part of the prompt tokens read from the provider cache. | [optional] [example: 19226] |
| cacheWriteTokens | Long (int64) | Part of the prompt tokens written to the provider cache. | [optional] [example: 104] |
| reasoningTokens | Long (int64) | Part of the completion tokens the model spent on reasoning. | [optional] [example: 68] |
| imageTokens | Long (int64) | Tokens spent on images. | [optional] [example: 0] |