Export the AI transaction log as CSV

get

https://api.revenium.ai/profitstream/v2/api/sources/metrics/ai/completions/export

Download every AI transaction matching the filters as a CSV file, oldest first, one row
per metered call. Takes the same filters as the list endpoint, so the window is an
arbitrary startDate/endDate pair at full ISO-8601 precision rather than a fixed preset.

The file carries the columns the transaction log shows — including trace, task type, agent
and subscriber — plus the provider and the individual token counts.

When the filters match more transactions than the export limit, or the file would be
larger than the export size limit, the request is rejected with 400 instead of returning
a shortened file.

Recent Requests

Time	Status	User Agent
Retrieving recent requests…

Loading…

Query Params

startDate

string

Filter start date/time in ISO 8601 format.

endDate

string

Filter end date/time in ISO 8601 format.

timeToFirstTokenMin

int64

0 to 60000

Minimum time to first token (in milliseconds).

timeToFirstTokenMax

int64

0 to 60000

Maximum time to first token (in milliseconds).

requestDurationMin

int64

0 to 300000

Minimum request duration (in milliseconds).

requestDurationMax

int64

0 to 300000

Maximum request duration (in milliseconds).

mediationLatencyMin

int64

0 to 10000

Minimum mediation latency (in milliseconds).

mediationLatencyMax

int64

0 to 10000

Maximum mediation latency (in milliseconds).

inputTokenCountMin

int64

0 to 10000000

Minimum input tokens.

inputTokenCountMax

int64

0 to 10000000

Maximum input tokens.

outputTokenCountMin

int64

0 to 10000000

Minimum output tokens.

outputTokenCountMax

int64

0 to 10000000

Maximum output tokens.

reasoningTokenCountMin

int64

0 to 5000000

Minimum reasoning tokens.

reasoningTokenCountMax

int64

0 to 5000000

Maximum reasoning tokens.

cachedTokenCountMin

int64

0 to 5000000

Minimum cached tokens.

cachedTokenCountMax

int64

0 to 5000000

Maximum cached tokens.

totalTokenCountMin

int64

0 to 20000000

Minimum total tokens.

totalTokenCountMax

int64

0 to 20000000

Maximum total tokens.

inputTokenCostMin

number

0 to 99999.99

Minimum input cost ($).

inputTokenCostMax

number

0 to 99999.99

Maximum input cost ($).

outputTokenCostMin

number

0 to 99999.99

Minimum output cost ($).

outputTokenCostMax

number

0 to 99999.99

Maximum output cost ($).

totalCostMin

number

0 to 999999.99

Minimum total cost ($).

totalCostMax

number

0 to 999999.99

Maximum total cost ($).

transactionId

string

length between 1 and 255

Exact transaction ID match.

traceId

string

length between 1 and 255

Exact trace ID match.

traceType

string

length between 1 and 255

Filter by trace type. Use 'unknown' for metrics without traceType.

traceName

string

length between 1 and 255

Filter by trace name.

organizationName

string

length between 1 and 255

Filter by organization name.

subscriberEmail

string

Filter by subscriber email.

subscriberId

string

length between 1 and 255

Filter by subscriber ID.

subscriptionId

string

length between 1 and 255

Filter by subscription ID.

productId

string

length between 1 and 255

Filter by product ID.

agent

string

length between 1 and 255

Filter by agent name.

provider

string

length between 1 and 255

Filter by AI provider.

model

string

length between 1 and 255

Filter by AI model.

modelSource

string

length between 1 and 255

Filter by model source.

taskType

string

length between 1 and 255

Filter by task type.

temperatureMin

double

0 to 2

Minimum temperature value.

temperatureMax

double

0 to 2

Maximum temperature value.

stopReason

string

enum

Filter by stop reason.

errorReason

string

length between 1 and 255

Filter by error reason.

responseQualityScoreMin

number

0 to 1

Minimum response quality score.

responseQualityScoreMax

number

0 to 1

Maximum response quality score.

credentialName

string

length between 1 and 255

Filter by credential name.

systemFingerprint

string

length between 1 and 255

Filter by system fingerprint.

environment

string

length between 1 and 255

Filter by environment (exact match).

includeCodingAssistants

boolean

Whether to include coding assistant metrics (e.g. Claude Code, Gemini CLI). When false (default), coding assistant records are excluded from results.

includeCodingAssistantProviders

array of strings

Server-side coding assistant provider pricing policy. Accepted values: ClaudeCode, ClaudeCowork, CursorIde, GeminiCli, CodexCli, GithubCopilot. Providers in this list are treated as API-rate real spend and included on real-cost surfaces. An empty or omitted list treats coding assistant providers as subscription or seat-based usage and excludes them from real-cost defaults.

includeCodingAssistantProviders

teamId

string

required

Unique identifier of the team (organization). Format: Revenium hashid (URL-safe encoded string).

Headers

string

enum

Defaults to application/json

Generated from available response content types

Allowed:

Responses

Language

Credentials

Header

Loading…

Response

Click Try It! to start a request and see the response here! Or choose an example:

application/json

*/*

200A CSV file of the matching AI transactions.

400Invalid filters, or the window exceeds the export's transaction or size limit

401Unauthorized - Authentication required

403Forbidden - user is not authorized to view the team

404team not found

409Conflict

422Unprocessable entity - Validation failed

429Too Many Requests

500Internal server error