> ## Documentation Index
> Fetch the complete documentation index at: https://platform.kimi.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Get usage aggregation

> Get usage aggregation

> [!NOTE]
>
> This endpoint requires an Admin Key with the `billing:read` scope.
>
> Window caps depend on granularity: `1m` up to 24 hours, `1h` up to 31 days, `1d` up to 93 days. `uid[]` is a filter dimension (only rows of keys owned by these users are returned); rows are grouped by bucket × model × project × key.

## Granularity, alignment, and empty buckets

| Granularity | Bucket alignment | Window cap per query | Empty buckets |
| - | - | - | - |
| `1m` | Rounded down to the minute (Asia/Shanghai) | ≤24 hours | No rows returned for buckets without data |
| `1h` | Rounded down to the hour (Asia/Shanghai) | ≤31 days | Same as above |
| `1d` | Calendar days (Asia/Shanghai) | ≤93 days | Same as above |

Data retention follows the platform retention policy (no fixed number of days is guaranteed); rows are grouped by bucket × model × project × key.


## OpenAPI

````yaml GET /v1/admin/organizations/usage/{usage_type}
openapi: 3.0.3
info:
  title: Kimi Admin API
  version: 2026-10-20-beta
  description: >-
    ## Overview


    The Kimi Admin API lets organization administrators manage members,
    invitations, projects, and inference API keys programmatically, and query
    billing, usage, and audit logs — without signing in to the console. It is a
    control-plane API: Admin API credentials cannot call any inference endpoint,
    and inference API keys cannot call this API.


    The base URL is `https://api.moonshot.ai/v1`.


    <Note>

    The Admin API is in **beta** and opens to organizations that have completed
    organization verification. After completing organization verification under
    **Console → Organization → Organization Verification**, organization admins
    can create Admin API keys under **Organization → Admin API Keys**.

    </Note>


    ## Authentication


    Send the Admin API secret (`sk-admin-*`) as a Bearer token:


    ```sh

    curl https://api.moonshot.ai/v1/admin/organizations \
      --header "Authorization: Bearer $MOONSHOT_ADMIN_KEY"
    ```


    A key is bound to exactly one organization; the organization is always
    derived from the key — cross-organization access is impossible by
    construction. If the key owner leaves the organization or is deactivated,
    the key stops working within 30 seconds. Secrets are shown only once, at
    creation or rotation; store them safely.


    ## Scopes


    Every key carries scopes chosen at creation:


    | Scope | Type | Covers |

    | --- | --- | --- |

    | `billing:read` | read | Daily/product bills, request details, interrupted
    requests, account balance, usage aggregation |

    | `org:read` | read | Members, invitations, projects, project members,
    inference API key metadata, spend limits, organization rate limits |

    | `org:write` | write | Member/invitation/project/key writes (implies
    `org:read` — no separate grant needed) |

    | `audit:read` | read | Audit log events |

    | `all` | read+write | Everything above; new scopes are covered
    automatically |


    Write endpoints additionally require the key owner to be an organization
    admin at call time and are rate-limited more strictly (see below).


    ## Rate limits


    | Class | Limit |

    | --- | --- |

    | Read endpoints | 60 requests/minute per organization |

    | Write endpoints | 20 requests/minute per organization |

    | Request details, interrupted requests, usage aggregation | 5 concurrent
    requests per organization |


    Rate limits are counted per organization and endpoint: all Admin Keys in the
    same organization share the quota — more keys do not raise it. Exceeding a
    limit returns `429 rate_limit_exceeded` / `concurrency_limit_exceeded`.


    ## Pagination


    Paginated endpoints (daily/product bills, request details, interrupted
    requests, usage aggregation, audit logs) use cursor pagination: responses
    carry `has_more` and `next`; pass `next` back as the `cursor` query
    parameter. Page size by endpoint: daily/product bills and usage default 100,
    max 1000; request details and interrupted requests default 100, max 10000;
    audit logs default 20, max 100. The cursor carries only the page position —
    restate all filters on every page.


    Organization collection endpoints (members, invitations, projects, project
    members, API keys) return the full list in one response, with no pagination
    parameters. Query parameters outside the protocol (such as `page_size`) are
    silently ignored — double-check parameter names.


    ## Errors


    Errors follow a unified envelope:


    ```json

    {
      "error": {
        "type": "invalid_request_error",
        "code": "invalid_date_range",
        "message": "end must be greater than start.",
        "param": "end",
        "request_id": "..."
      }
    }

    ```


    `data_not_ready` gates the bill endpoints' readiness: current-day data
    becomes ready on T-1 and is queryable from 07:00 CST the next day.


    ## Data conventions


    - Monetary amounts are in **milli-cent** (1 USD = 100,000 milli-cent).

    - Timestamps in query parameters are epoch **milliseconds**; timestamps in
    response bodies are RFC 3339 unless stated otherwise.

    - Time windows are left-closed, right-open, aligned to CST day boundaries;
    timestamps in query parameters are epoch milliseconds, and RFC 3339 in
    responses unless stated otherwise. Passing seconds returns empty data rather
    than an error — check your units.

    - Billing data becomes ready on T-1: the current day is queryable from 07:00
    CST the next day and returns `400 data_not_ready` earlier; interrupted
    request details are exempt (near real time).

    - Retention and window caps: daily/product bills 365 days; request details
    and interrupted requests 90 days with a 31-day per-call cap; audit logs 31
    days per call; usage aggregation windows by granularity (`1m` ≤24 hours,
    `1h` ≤31 days, `1d` ≤93 days, cut on Asia/Shanghai natural days).

    - Each organization holds at most 10 active Admin API keys.
servers:
  - url: https://api.moonshot.ai
    description: Production
security:
  - AdminBearerAuth: []
paths:
  /v1/admin/organizations/usage/{usage_type}:
    get:
      tags:
        - Usage
      summary: Get usage aggregation
      operationId: getUsage
      parameters:
        - name: usage_type
          in: path
          required: true
          schema:
            type: string
            enum:
              - completions
          description: Usage type. Only completions is supported currently.
        - $ref: '#/components/parameters/Start'
        - $ref: '#/components/parameters/End'
        - $ref: '#/components/parameters/ProjectID'
        - $ref: '#/components/parameters/APIKey'
        - name: uid
          in: query
          required: false
          schema:
            type: array
            items:
              type: string
          style: form
          explode: true
          description: >-
            User ID filters. Rows are restricted to these users within the
            organization.
        - name: granularity
          in: query
          required: false
          schema:
            type: string
            enum:
              - 1m
              - 1h
              - 1d
            default: 1d
          description: >-
            Time bucket granularity. Window caps differ by granularity: 1m
            allows up to 24 hours, 1h up to 31 days, 1d up to 93 days; `1d`
            buckets are cut on Asia/Shanghai natural days.
        - name: limit
          in: query
          required: false
          schema:
            type: integer
            minimum: 1
            maximum: 1000
            default: 100
          description: Page size. Defaults to 100, maximum 1000.
        - name: cursor
          in: query
          required: false
          schema:
            type: string
          description: >-
            Cursor from the previous response `next` field. The cursor carries
            only the page position; all query parameters must be restated on
            every page. Keep them unchanged while paginating (except limit) —
            the server does not check them for consistency.
      responses:
        '200':
          description: Usage aggregation rows, ascending by bucket time.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/UsageResponse'
        '400':
          description: >-
            Invalid request parameters (illegal/out-of-retention time window,
            invalid enum value, or illegal limit/cursor).
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/BillingErrorResponse'
        '401':
          description: >-
            Authentication failed (the Admin API key is invalid,
            deleted/expired, or its owner is no longer active).
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/BillingErrorResponse'
        '403':
          description: Insufficient scope (the Admin Key lacks the billing:read scope).
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/BillingErrorResponse'
        '404':
          description: Resource not found (no data for this organization).
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/BillingErrorResponse'
        '429':
          description: >-
            Rate limit exceeded (requests-per-minute cap or concurrency cap
            hit).
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/BillingErrorResponse'
        '500':
          description: Internal server error (contact support with the request_id).
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/BillingErrorResponse'
      security:
        - AdminBearerAuth: []
components:
  parameters:
    Start:
      name: start
      in: query
      required: true
      schema:
        type: integer
        format: int64
        minimum: 0
      description: >-
        Start time in milliseconds, inclusive, e.g. `1790784000000` (2026-10-01
        00:00 CST).
      example: 1790784000000
    End:
      name: end
      in: query
      required: true
      schema:
        type: integer
        format: int64
        minimum: 0
      description: >-
        End time in milliseconds, exclusive, e.g. `1792598400000` (2026-10-22
        00:00 CST).
      example: 1792598400000
    ProjectID:
      name: project_id
      in: query
      required: false
      schema:
        type: array
        items:
          type: string
      style: form
      explode: true
      description: >-
        Project IDs. Every project must belong to the authenticated
        organization.
    APIKey:
      name: key
      in: query
      required: false
      schema:
        type: array
        items:
          type: string
      style: form
      explode: true
      description: >-
        Business API key IDs. Every key must belong to the authenticated
        organization.
  schemas:
    UsageResponse:
      type: object
      required:
        - data
        - has_more
        - next
      additionalProperties: false
      properties:
        data:
          type: array
          description: Usage aggregation rows (bucket × model × project × API key).
          items:
            $ref: '#/components/schemas/UsageBucket'
        has_more:
          type: boolean
          description: Whether there is a next page.
        next:
          type: string
          nullable: true
          description: >-
            Cursor for the next page. Pass it back verbatim as the `cursor`
            query parameter.
    BillingErrorResponse:
      type: object
      required:
        - error
      additionalProperties: false
      properties:
        error:
          $ref: '#/components/schemas/BillingErrorBody'
    UsageBucket:
      type: object
      required:
        - bucket_time
        - model_id
        - project_id
        - key
        - prompt_tokens
        - completion_tokens
        - cached_tokens
        - request_count
      additionalProperties: false
      properties:
        bucket_time:
          type: string
          format: date-time
          description: >-
            Bucket start time (RFC3339, granularity-aligned; 1d buckets align to
            Asia/Shanghai natural days).
        model_id:
          type: string
          description: Model ID.
        project_id:
          type: string
          description: Project ID.
        key:
          type: string
          description: Business API key ID.
        prompt_tokens:
          type: integer
          format: int64
          minimum: 0
          description: Input tokens in this bucket.
        completion_tokens:
          type: integer
          format: int64
          minimum: 0
          description: Output tokens in this bucket.
        cached_tokens:
          type: integer
          format: int64
          minimum: 0
          description: >-
            Cache-hit tokens in this bucket (included in prompt_tokens; billable
            prompt = prompt_tokens − cached_tokens. Metered separately from
            cache_write_* (cache creation), so no double counting).
        cache_write_5m_tokens:
          type: integer
          format: int64
          minimum: 0
          description: >-
            Cache write tokens (5-minute TTL tier) summed over this bucket.
            Omitted when every row in the bucket predates tiered cache-write
            metering (legacy rows); field absent, not null — same semantics as
            request details.
        cache_write_1h_tokens:
          type: integer
          format: int64
          minimum: 0
          description: >-
            Cache write tokens (1-hour TTL tier) summed over this bucket.
            Omitted when every row in the bucket predates tiered cache-write
            metering (legacy rows); field absent, not null — same semantics as
            request details.
        request_count:
          type: integer
          format: int64
          description: Number of requests in this bucket.
    BillingErrorBody:
      type: object
      required:
        - type
        - code
        - message
        - request_id
      additionalProperties: false
      properties:
        type:
          type: string
          example: invalid_request_error
        code:
          type: string
          description: Billing-domain error codes (subset).
          enum:
            - invalid_request
            - invalid_date_range
            - date_range_too_large
            - date_range_out_of_retention
            - data_not_ready
            - invalid_cursor
            - invalid_api_key
            - api_key_inactive
            - insufficient_scope
            - resource_not_found
            - rate_limit_exceeded
            - concurrency_limit_exceeded
            - internal_server_error
        message:
          type: string
        param:
          type: string
          nullable: true
        request_id:
          type: string
  securitySchemes:
    AdminBearerAuth:
      type: http
      scheme: bearer
      bearerFormat: Admin API Secret (sk-admin-*)
      description: >-
        Send the Admin API secret as a Bearer token in the Authorization header
        (`Bearer sk-admin-***`). Create Admin Keys under Console → Organization
        → Admin API Keys.

````