curl --request POST \
--url https://api.select.dev/v2/metrics/bigquery-all-compute/query \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--header 'x-tenant-id: <x-tenant-id>' \
--data '
{
"measures": [
"cost",
"saved_vs_on_demand",
"utilization_efficiency_percent"
],
"dimensions": [
"reservation_id",
"reservation_label",
"reservation_type"
],
"filter_expression": {
"operator": "and",
"filters": [
{
"field": "day",
"operator": "in range",
"value": "30D"
}
]
},
"sort": [
{
"field": "cost",
"direction": "desc"
}
],
"limit": 50
}
'{
"items": [
{}
],
"page_token": "<string>",
"row_count": 123,
"sql": "<string>"
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}Query BigQuery compute by reservation
One row per group of BigQuery jobs, by the reservation that ran them or by on-demand pricing, carrying job cost and usage alongside the reservation’s capacity, utilization and cost.
One filter limit applies. When the request selects or filters on a reservation attribute (reservation_label, reservation_group_label, reservation_name, commitment_plan, renewal_plan, is_flat_rate, bigquery_reservation_labels_map, bigquery_reservation_labels) or a measure that reads reservation capacity, such as cost, reserved_slot_s or utilization_efficiency_percent, each filter on a time field (hour, day, week, month, quarter) or on a reservation attribute also applies on its own, as if it were joined with and, even inside an or group. Do not put these fields under or.
Returns one row per combination of the requested dimensions, carrying the requested measures. Each row is keyed by dimension and measure name. row_count is the number of rows in items: results are not paginated, so an aggregation with more rows than limit is cut short at limit rather than continued under a page_token. Send X-Team-Id to evaluate the query as a team you belong to; rows are then limited to what that team’s roles reach.
Use this route for the billed cost of BigQuery compute by reservation, with on-demand jobs beside the reservations, and for how well each reservation used its capacity. For job cost by user, project or workload only, use bigquery-workloads, which reads less. For slot capacity over time without cost, use bigquery-reservation-timeline. cost is the billed cost: for a reservation, its reserved capacity at the rate of each capacity tier, used or idle; for on-demand jobs, the bytes billed at the on-demand rate. sum_spend is the job cost attributed from the workloads, so the two do not have to agree. saved_vs_on_demand is the on-demand cost of the same bytes minus cost. Naming hour, start_hour or start_hour_of_day reads the hourly table. All other requests read the daily table. Do not put time fields or reservation attributes under or.
curl --request POST \
--url https://api.select.dev/v2/metrics/bigquery-all-compute/query \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--header 'x-tenant-id: <x-tenant-id>' \
--data '
{
"measures": [
"cost",
"saved_vs_on_demand",
"utilization_efficiency_percent"
],
"dimensions": [
"reservation_id",
"reservation_label",
"reservation_type"
],
"filter_expression": {
"operator": "and",
"filters": [
{
"field": "day",
"operator": "in range",
"value": "30D"
}
]
},
"sort": [
{
"field": "cost",
"direction": "desc"
}
],
"limit": 50
}
'{
"items": [
{}
],
"page_token": "<string>",
"row_count": 123,
"sql": "<string>"
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}{
"title": "<string>",
"status": 123,
"detail": "<string>",
"code": "validation_failed",
"retryable": true,
"type": "about:blank",
"details": [
{}
]
}Authorizations
Organization API key (sl_…).
Headers
Act as this team: rows are limited to what the team's roles reach instead of the caller's own. The caller must be a member of the team or hold the organization-wide permission the route requires.
The organization ID the request is scoped to.
Body
Aggregate BigQuery compute by reservation, with on-demand jobs beside them.
A row describes BigQuery jobs grouped by the reservation that ran them, or
by on-demand pricing, with the reservation's capacity, utilization and cost
joined in. Daily and hourly rollups answer this model. A request that names
hour, start_hour or start_hour_of_day reads the hourly rollup;
everything else reads the cheaper daily one.
Categorize against these set definitions instead of the organization's saved usage group sets, for evaluating a set before saving it. Omit to use the saved sets, which is what every other surface reports against.
Show child attributes
Show child attributes
Aggregates to compute over the matching rows.
Show child attributes
Show child attributes
Columns to group the aggregates by. Each is returned on every row.
Show child attributes
Show child attributes
An and/or tree of predicates over dimensions, applied before aggregation. Omit to match every row the caller can see.
Show child attributes
Show child attributes
Predicates over measures, applied after aggregation.
- InFilter[OrgBigQueryHourlyReservationMeasures]
- ArrayContainsFilter[OrgBigQueryHourlyReservationMeasures]
- IsFilter[OrgBigQueryHourlyReservationMeasures]
- LikeFilter[OrgBigQueryHourlyReservationMeasures]
- MultiLikeFilter[OrgBigQueryHourlyReservationMeasures]
- EqFilter[OrgBigQueryHourlyReservationMeasures]
- FilterExpression[OrgBigQueryHourlyReservationMeasures]
- InRelativeDateRangeFilter[OrgBigQueryHourlyReservationMeasures]
Show child attributes
Show child attributes
Ordering applied to the aggregated rows.
Show child attributes
Show child attributes
Maximum rows to return. Omit for no limit. An aggregation that produces more rows than this is silently cut short at the limit.
x >= 1Return the SQL that produced the rows alongside them, for inspecting how a result was computed. Not every query endpoint publishes its SQL; where it does not, the field stays null.
Reshape the result so the named dimensions become columns rather than rows. Each remaining dimension yields one row, carrying a pivot_data object keyed by the pivoted dimension values. Every name in pivot.on must also appear in dimensions.
Show child attributes
Show child attributes
Return the result as a tree rather than a flat list, grouping the named dimensions into one level per tree depth. One request fetches the top level; a follow-up naming rollup.parent_paths fetches the children of the nodes expanded so far, so a deep hierarchy is read a level at a time. Every name in rollup.dimensions must also appear in dimensions.
Show child attributes
Show child attributes
Aggregate the matching rows. Set false to return the underlying rows as they are, projecting the requested dimensions with no grouping — a listing rather than an aggregation. It is honored only for a request that asks for no measures and no measure_filters, since an aggregate has to group to be computable; limit, when set, bounds the rows returned either way.
Response
Successful Response
Opaque cursor for the next page; empty/absent on the last page.
Best-effort total for the filtered result set; may be null when expensive.
The SQL that produced these rows, returned only when the request asks for it. Database and schema names are left as placeholders. Null when the request did not ask for it, or when this endpoint does not publish its SQL.