Get Account Cost ​

Returns the Cost and Provider Cost shown under Analysis > Cost in the OpenQiao console.

GET https://openqiao.com/api/v1/management/cost

Authentication ​

Authorization Header * ​

Authorization: Bearer <OPENQIAO_MANAGEMENT_API_KEY>

• Name: Authorization
• Format: Bearer <API_KEY>
• Description: A Management API Key created using your personal account in the OpenQiao Console

Query parameters ​

For type=cost, omitting api_key_ids, model_slugs, and bill_types (or leaving them empty) returns all summary and grouped statistics for the personal account that owns the Management Key, within the period specified by query_dimension and query_time.

For type=provider_cost, model_slugs is still required and must contain exactly one model. api_key_ids and bill_types can be omitted.

The dimension and time formats exactly match the console: month/day/hour queries use BIZ_MTH + YYYYMM, BIZ_DT + YYYYMMDD, and BIZ_HOUR + YYYYMMDDHH. Hour values are UTC.

api_key_ids only accepts IDs of non-deleted inference or subscription API Keys belonging to the authenticated personal account. Passing a Management Key ID, another account's Key ID, or a deleted Key ID returns HTTP 403.

Returns ​

success boolean ​

Whether the request succeeded. Always true in a successful response.

data object ​

Query result. Both cost and provider_cost contain summary (cost totals) and analysis (grouped analysis). The shape of analysis depends on type.

Non-null amounts, token counts, counts, ratios, and performance metrics are decimal strings, not JSON numbers. null means no statistic is available or the field is not used for that grouping; it is not equivalent to "0". A successful response may contain empty arrays and null summary fields when no data is available.

Time series use the following bizTime formats according to query_dimension:

data.summary object ​

Cost summary for the selected filters. Both cost and provider_cost use the same structure:

• totalCost string | null — Total cost for the selected filters, in USD.
• inputCost string | null — Input token cost, in USD.
• outputCost string | null — Output token cost, in USD.
• otherCost string | null — Other costs not classified as input or output token costs, in USD.
• requestCounts string | null — Total request count for the selected filters, not a monetary amount.
• requestAvgCost string | null — Average cost per request, in USD/request.
• totalTokens string | null — Total token count for the selected filters, not a monetary amount.
• millionTokenAvgCost string | null — Average cost normalized to one million tokens, in USD/million tokens, not the model’s list price.

Cost (type=cost) ​

data.analysis object ​

Contains the three cost time series below. They are different breakdowns of the same costs; do not add the amounts across these arrays.

data.analysis.costByModel array ​

Grouped by time and model. Each item contains:

• bizTime string — Time bucket identifier in the format listed above, not an individual request timestamp.
• modelSlug string — Model identifier, such as openai/gpt-5, not its display name.
• tokenType string | null — Not a token-type grouping; normally null.
• apiKeyId string | null — Not an API Key grouping; normally null.
• billAmount string | null — Cost for this group and time bucket, in USD.
• requestCounts string | null — Request count for this group and time bucket.

data.analysis.costByApiKey array ​

Grouped by time and API Key. Each item contains:

• bizTime string — Time bucket identifier in the format listed above, not an individual request timestamp.
• modelSlug string | null — Not a model grouping; normally null.
• tokenType string | null — Not a token-type grouping; normally null.
• apiKeyId string — API Key resource ID accepted by api_key_ids, not the secret key token.
• billAmount string | null — Cost for this group and time bucket, in USD.
• requestCounts string | null — Request count for this group and time bucket.

data.analysis.costByTokenType array ​

Grouped by time and token type. Each item contains:

• bizTime string — Time bucket identifier in the format listed above, not an individual request timestamp.
• modelSlug string | null — Not a model grouping; normally null.
• tokenType string — Billable token type, such as prompt (input) or completion (output).
• apiKeyId string | null — Not an API Key grouping; normally null.
• billAmount string | null — Cost for this group and time bucket, in USD.
• requestCounts string | null — Request counts are normally not returned for this grouping and are null.

Provider Cost (type=provider_cost) ​

model_slugs requires exactly one model. data.summary uses the cost summary fields above.

data.analysis object ​

Endpoint statistics for the selected model, with the fields below.

data.analysis.fastestEndpointSlug string | null ​

Identifier of the fastest endpoint within the query; may be null when unavailable.

data.analysis.mainEndpointSlug string | null ​

Identifier of the main endpoint with the most requests within the query; may be null when unavailable.

data.analysis.uptime string | null ​

Overall request success rate for the selected model, from 0 to 1. For example, "0.9920" means 99.20%; this is not elapsed uptime.

data.analysis.endpointCount string | null ​

Number of endpoints within the query, represented as a string.

data.analysis.endpointUsages array ​

Statistics grouped by endpoint. Each item contains:

• endpointSlug string — Endpoint identifier distinguishing the endpoints used for the selected model.
• requestCounts string | null — Total requests for this endpoint.
• tokens string | null — Total tokens for this endpoint.
• requestRatio string | null — Share of requests served by this endpoint within the query, from 0 to 1; "0.8" means 80%.
• tokenRatio string | null — Share of tokens served by this endpoint within the query, from 0 to 1.
• medianLatency string | null — Median request latency for this endpoint, in milliseconds (ms), not mean latency.
• medianThroughput string | null — Median throughput for this endpoint, in tokens per second.
• billAmount string | null — Cost for this endpoint, in USD.
• billAmountRatio string | null — Share of the query’s total cost attributable to this endpoint, from 0 to 1.

curl -G https://openqiao.com/api/v1/management/cost \
  -H "Authorization: Bearer $OPENQIAO_MANAGEMENT_API_KEY" \
  --data-urlencode "type=cost" \
  --data-urlencode "query_dimension=BIZ_MTH" \
  --data-urlencode "query_time=202609" \
  --data-urlencode "api_key_ids=key_1,key_2" \
  --data-urlencode "model_slugs=openai/gpt-5" \
  --data-urlencode "bill_types=metered,subscription"

{
  "success": true,
  "data": {
    "summary": {
      "totalCost": "12.3456000000",
      "inputCost": "8.0000000000",
      "outputCost": "4.3456000000",
      "otherCost": "0",
      "requestCounts": "125",
      "requestAvgCost": "0.0987648000",
      "totalTokens": "1500000",
      "millionTokenAvgCost": "8.2304000000"
    },
    "analysis": {
      "costByModel": [
        {
          "bizTime": "20260903",
          "modelSlug": "openai/gpt-5",
          "tokenType": null,
          "apiKeyId": null,
          "billAmount": "12.3456000000",
          "requestCounts": "125"
        }
      ],
      "costByApiKey": [
        {
          "bizTime": "20260903",
          "modelSlug": null,
          "tokenType": null,
          "apiKeyId": "key_1",
          "billAmount": "8.0000000000",
          "requestCounts": "80"
        },
        {
          "bizTime": "20260903",
          "modelSlug": null,
          "tokenType": null,
          "apiKeyId": "key_2",
          "billAmount": "4.3456000000",
          "requestCounts": "45"
        }
      ],
      "costByTokenType": [
        {
          "bizTime": "20260903",
          "modelSlug": null,
          "tokenType": "prompt",
          "apiKeyId": null,
          "billAmount": "8.0000000000",
          "requestCounts": null
        },
        {
          "bizTime": "20260903",
          "modelSlug": null,
          "tokenType": "completion",
          "apiKeyId": null,
          "billAmount": "4.3456000000",
          "requestCounts": null
        }
      ]
    }
  }
}

curl -G https://openqiao.com/api/v1/management/cost \
  -H "Authorization: Bearer $OPENQIAO_MANAGEMENT_API_KEY" \
  --data-urlencode "type=provider_cost" \
  --data-urlencode "query_dimension=BIZ_DT" \
  --data-urlencode "query_time=20260903" \
  --data-urlencode "model_slugs=openai/gpt-5"

{
  "success": true,
  "data": {
    "summary": {
      "totalCost": "12.3456000000",
      "inputCost": "8.0000000000",
      "outputCost": "4.3456000000",
      "otherCost": "0",
      "requestCounts": "125",
      "requestAvgCost": "0.0987648000",
      "totalTokens": "1500000",
      "millionTokenAvgCost": "8.2304000000"
    },
    "analysis": {
      "fastestEndpointSlug": "provider-a",
      "mainEndpointSlug": "provider-a",
      "uptime": "0.9990",
      "endpointCount": "2",
      "endpointUsages": [
        {
          "endpointSlug": "provider-a",
          "requestCounts": "100",
          "tokens": "1200000",
          "billAmount": "9.1234000000",
          "billAmountRatio": "0.7391"
        }
      ]
    }
  }
}

Rate limits and errors ​

Account Usage and Cost endpoints share a limit of 60 requests per minute for each personal account. Request 61 returns HTTP 429 and Retry-After: 60. Existing platform-wide Management API limits also apply.