LLM and Generative AI Usage Logs API

Back to API List This API retrieves LLM and Generative AI usage logs for GenAI features.

Field	Value
Method	POST
Endpoint	`https://{{host}}/api/1.1/public/bot/{{botId}}/getLLMUsageLogs?version=2.0`
Content Type	`application/json`
Authorization	`auth: {{JWT}}` — See How to generate the JWT Token.
API Scope	App Builder: Fetch Gen AI and LLM Usage Logs; Admin Console: API Scopes > Gen AI and LLM Usage Logs

A single API call supports querying up to 90 days of data and retrieving up to 50 records. Make multiple requests using skip and limit to retrieve the complete dataset.

Path Parameters

Parameter	Required/Optional	Description
`host`	Required	The environment URL. For example, `https://platform.kore.ai`
`botId`	Required	Bot identifier or Stream identifier. Accessible from the General Settings page of the bot.

Query Parameters

Parameter	Required/Optional	Description
`version`	Optional	API version number. For support backward compatibility, the request must include the `version=2.0` in the query parameter.

Sample Request version 1.0

curl --location 'https://{{host}}/api/1.1/public/bot/{{botId}}/getLLMUsageLogs?' \
--header 'auth: {{JWT_TOKEN}}'\
--header 'Content-Type: application/json' \
--data '{
    "dateFrom": "2024-03-07",
    "dateTo": "2024-04-26",
    "limit": "50",
    "skip": "5",
    "isDeveloper": true,
    "channel": ["msteams", "rtm"],
    "featureName": ["Agent Node"],
    "taskId": "dg-d4924db4-xxxx-5b4d-ae51-aa5c7be4xxxx",
    "channelUserIds": ["29:1gmwl6k7ohtdhxxxx7kvu1yiek_qri4grpmd0k_am9dpzwttw"],
    "userIds": ["u-40b3eafc-xxxx-55f2-83e8-cf4d0fbxxxx"],
    "sort": {
        "field": "Time Taken",
        "order": "asc"
    }
}'

Sample Request version 2.0

curl --location 'https://{{host}}/api/1.1/public/bot/{{botId}}/getLLMUsageLogs?version=2.0' \
--header 'auth: {{JWT_TOKEN}}'\
--header 'Content-Type: application/json' \
--data '{
    "dateFrom": "2024-03-07",
    "dateTo": "2024-04-26",
    "limit": "50",
    "skip": "5",
    "isDeveloper": true,
    "channel": ["msteams", "rtm"],
    "featureName": ["Agent Node"],
    "taskId": "dg-d4924db4-xxxx-5b4d-ae51-aa5c7be4xxxx",
    "channelUserIds": ["29:1gmwl6k7ohtdhxxxx7kvu1yiek_qri4grpmd0k_am9dpzwttw"],
    "userIds": ["u-40b3eafc-xxxx-55f2-83e8-cf4d0fbxxxx"],
    "sort": {
        "field": "Time Taken",
        "order": "asc"
    }
}'

Body Parameters

Parameter	Type	Required	Description
`dateFrom`	String	Yes	Start date for retrieving LLM usage logs. Format: `YYYY-MM-DD`
`dateTo`	String	Yes	End date for retrieving LLM usage logs. Format: `YYYY-MM-DD`
`limit`	Integer	Yes	Maximum number of records to return.
`skip`	Integer	No	Number of records to skip before returning results. Used for pagination.
`isDeveloper`	Boolean	No	Filters logs generated by developer interactions when set to `true`.
`channel`	Array	No	List of channels to filter the logs. Example: `["msteams", "rtm"]`.
`featureName`	Array	No	List of feature names to filter the logs. Example: `["Agent Node"]`.
`taskId`	String	No	Unique identifier of the task associated with the usage logs.
`channelUserIds`	Array	No	List of channel-specific user IDs used to filter logs.
`userIds`	Array	No	List of platform user IDs used to filter logs.
`sort`	Object	No	Defines the sorting criteria for the results.
`sort.field`	String	No	Field used for sorting the results. Example: `Time Taken`.
`sort.order`	String	No	Sort order. Supported values: `asc`, `desc`.

Sample Response version 1.0

[
    {
      "promptName": "Default",
      "integration": "OpenAI",
      "startDate": "2024-03-08T06:38:54.008Z",
      "endDate": "2024-03-08T06:38:58.386Z",
      "timeTaken": 4378,
      "statusCode": 200,
      "botId": "st-73bfdb2f-xxxx-55e6-b4c8-3f568a6exxxx",
      "userId": "u-40b3eafc-xxxx-55f2-83e8-cf4d0fb1xxxx",
      "featureName": "GenAI Node",
      "modelName": "GPT-4",
      "channelName": "msteams",
      "description": "Order pizza-GenAINode0002",
      "taskId": "dg-200ad1ff-xxxx-5219-810e-4ee3800dxxxx",
      "status": "Success",
      "payloadDetails": {
        "requestPayload": {
          "model": "gpt-4",
          "temperature": 0.5,
          "maxTokens": 1500,
          "topP": 1,
          "frequencyPenalty": 0,
          "presencePenalty": 0,
          "messages": [
            {
              "role": "system",
              "content": "You are a virtual assistant..."
            },
            {
              "role": "user",
              "content": "Order pizza"
            }
          ]
        },
        "responsePayload": {
          "id": "chatcmpl-90OAYDFFkbjgMxhMxKukwqklT6SWO",
          "object": "chat.completion",
          "created": 1709879934,
          "model": "gpt-4-0613",
          "choices": [
            {
              "index": 0,
              "message": {
                "role": "assistant",
                "content": "{\"bot\":\"I'm sorry, but your message seems incomplete...\"}"
              },
              "finishReason": "stop"
            }
          ],
          "usage": {
            "promptTokens": 614,
            "completionTokens": 74,
            "totalTokens": 688
          }
        },
        "requestTokens": 614,
        "responseTokens": 74
      }
    }
  ]

Sample Response version 2.0

{
  "moreAvailable": false,
  "result": [
    {
      "promptName": "Default",
      "integration": "OpenAI",
      "startDate": "2024-03-08T06:38:54.008Z",
      "endDate": "2024-03-08T06:38:58.386Z",
      "timeTaken": 4378,
      "statusCode": 200,
      "botId": "st-73bfdb2f-xxxx-55e6-b4c8-3f568a6exxxx",
      "userId": "u-40b3eafc-xxxx-55f2-83e8-cf4d0fb1xxxx",
      "featureName": "GenAI Node",
      "modelName": "GPT-4",
      "channelName": "msteams",
      "description": "Order pizza-GenAINode0002",
      "taskId": "dg-200ad1ff-xxxx-5219-810e-4ee3800dxxxx",
      "status": "Success",
      "payloadDetails": {
        "requestPayload": {
          "model": "gpt-4",
          "temperature": 0.5,
          "maxTokens": 1500,
          "topP": 1,
          "frequencyPenalty": 0,
          "presencePenalty": 0,
          "messages": [
            {
              "role": "system",
              "content": "You are a virtual assistant..."
            },
            {
              "role": "user",
              "content": "Order pizza"
            }
          ]
        },
        "responsePayload": {
          "id": "chatcmpl-90OAYDFFkbjgMxhMxKukwqklT6SWO",
          "object": "chat.completion",
          "created": 1709879934,
          "model": "gpt-4-0613",
          "choices": [
            {
              "index": 0,
              "message": {
                "role": "assistant",
                "content": "{\"bot\":\"I'm sorry, but your message seems incomplete...\"}"
              },
              "finishReason": "stop"
            }
          ],
          "usage": {
            "promptTokens": 614,
            "completionTokens": 74,
            "totalTokens": 688
          }
        },
        "requestTokens": 614,
        "responseTokens": 74
      }
    }
  ],
  "total": 6
}

Response Parameters

Field	Type	Description
`moreAvailable`	Boolean	Indicates whether more records are available. `true`: Indicates records are available beyond the current response. `false`: No additional records are available.
`result`	Array	Contains the LLM usage log records.
`promptName`	String	Name of the prompt used for the request.
`integration`	String	Name of the LLM integration used for the request.
`startDate`	String	Start time of the request in ISO 8601 format.
`endDate`	String	End time of the request in ISO 8601 format.
`timeTaken`	Integer	Time taken to process the request, in milliseconds.
`statusCode`	Integer	HTTP status code returned for the request.
`botId`	String	Unique identifier of the bot.
`userId`	String	Unique identifier of the user.
`featureName`	String	Name of the feature that triggered the request.
`modelName`	String	Name of the LLM model used for the request.
`channelName`	String	Name of the channel from which the request originated.
`description`	String	Description of the request.
`taskId`	String	Unique identifier of the task.
`status`	String	Execution status of the request.
`payloadDetails`	Object	Contains request and response payload details.
`requestPayload`	Object	Contains the payload sent to the LLM provider.
`model`	String	Name of the model used for the request.
`temperature`	Number	Sampling temperature used to generate the response.
`maxTokens`	Integer	Maximum number of tokens that can be generated in the response.
`topP`	Number	Nucleus sampling value used for response generation.
`frequencyPenalty`	Number	Penalty applied to reduce repeated tokens.
`presencePenalty`	Number	Penalty applied to encourage new topics in the response.
`messages`	Array	Contains the messages sent in the request.
`role`	String	Role associated with the message, such as `system` or `user`.
`content`	String	Content of the message.
`responsePayload`	Object	Contains the response returned by the LLM provider.
`id`	String	Unique identifier of the response.
`object`	String	Type of response object.
`created`	Integer	Unix timestamp indicating when the response was created.
`choices`	Array	Contains the generated response choices.
`index`	Integer	Index of the response choice.
`message`	Object	Contains the generated message.
`finishReason`	String	Reason the model stopped generating the response.
`usage`	Object	Contains token usage information.
`promptTokens`	Integer	Number of tokens used in the prompt.
`completionTokens`	Integer	Number of tokens generated in the response.
`totalTokens`	Integer	Total number of tokens used for the request and response.
`requestTokens`	Integer	Number of tokens in the request.
`responseTokens`	Integer	Number of tokens in the response.
`total`	Integer	Total number of records returned or matching the query.

​Path Parameters

​Query Parameters

​Sample Request version 1.0

​Sample Request version 2.0

​Body Parameters

​Sample Response version 1.0

​Sample Response version 2.0

​Response Parameters

Path Parameters

Query Parameters

Sample Request version 1.0

Sample Request version 2.0

Body Parameters

Sample Response version 1.0

Sample Response version 2.0

Response Parameters