---
title: "Create Tracking Categories in batch"
method: POST
path: "/accounting/tracking-categories/batch"
tags: ["Tracking Categories"]
---

# Create Tracking Categories in batch

`POST /accounting/tracking-categories/batch`

## Request

- Base URL: `https://unify.apideck.com`
- URL: `https://unify.apideck.com/accounting/tracking-categories/batch`
- Auth: API key in header `Authorization`

Create multiple tracking categories in a single request. Each item is processed independently, so some may be created while others fail; the response returns `200` with one result per item, in the order the items were sent. The maximum number of items accepted per request depends on the connector.

## Query parameters

- `raw` boolean

## Headers

- `x-apideck-consumer-id` string, required
- `x-apideck-app-id` string, required
- `x-apideck-service-id` string
- `x-apideck-company-id` string

## Request body

- BatchTrackingCategoriesRequest: A batch of tracking categories to write in a single request. Each item is processed independently: some may succeed while others fail, and the response carries one result per item in the order they were sent.
  - `items` object[], required: The records to write. The per-connector limit is the real cap and is usually lower than the ceiling here: read `batch_support.resources[<resource>].max_items` on the Connector API for the connector you are calling, or the resource gotchas. This ceiling exists so an oversized array is rejected by request validation before any per-item work runs, rather than after it.
    - `_ref` string: A caller-supplied reference for this item, echoed back on the matching result so each outcome can be tied to the record it came from. Must be unique within the request. Never sent to the connector.
    - `data` TrackingCategoryCreateInput, required: The writable shape of TrackingCategory for a batch create. Identical to [TrackingCategory](the model returned on reads) with the read-only properties removed — the same properties the single-record create endpoint rejects, so a batched record and a single-record create accept exactly the same body.
      - `parent_id` string, nullable: A unique identifier for an object.
      - `parent_name` string, nullable: The name of the parent tracking category.
      - `name` string: The name of the tracking category.
      - `code` string, nullable: The code of the tracking category.
      - `status` 'active' | 'inactive': Based on the status some functionality is enabled or disabled.
      - `row_version` string, nullable: A binary value used to detect updates to a object and prevent data conflicts. It is incremented each time an update is made to the object.
      - `pass_through` object[]: The pass_through property allows passing service-specific, custom data or structured modifications in request body when creating or updating resources.
        - `service_id` string, required: Identifier for the service to which this pass_through should be applied.
        - `operation_id` string: Optional identifier for a workflow operation to which this pass_through should be applied. This is useful for Unify calls that are making more than one downstream request.
        - `extend_object` object: Simple object allowing any properties for direct extension.
        - `extend_paths` object[]: Array of objects for structured data modifications via paths.
          - `path` string, required: JSONPath string specifying where to apply the value.
          - `value` unknown, required
      - `subsidiaries` object[]: The subsidiaries the account belongs to.
        - `id` string: The ID of the subsidiary.

## Response `200`

Tracking Categories batch processed

- BatchTrackingCategoriesResponse
  - `status_code` integer, required: HTTP Response Status Code
  - `status` string, required: HTTP Response Status
  - `service` string, required: Apideck ID of service provider
  - `resource` string, required: Unified API resource name
  - `operation` string, required: Operation performed
  - `data` BatchItemResult[], required: One result per item in the request, in the order the items were sent.
    - `_ref` string: A caller-supplied reference for this item, echoed back on the matching result so each outcome can be tied to the record it came from. Must be unique within the request. Never sent to the connector.
    - `id` string: A unique identifier for an object.
    - `downstream_id` string, nullable: The third-party API ID of original entity
    - `status` 'created' | 'updated' | 'failed' | 'uncertain', required: The outcome for this item. - `created` / `updated` — the record was written. - `failed` — it was not written, and `error` says why. - `uncertain` — the request to the connector did not complete (typically a timeout), so whether this record was written is genuinely unknown. Do not blindly retry: list the resource to find what exists first, or a retry may duplicate it. A `created` or `updated` result can also carry an `error`: the record was written, but work that runs after the write did not complete. Read this rather than `id` to decide what happened. `id` is present when the identifier could also be read back, which is not the same question — a written record whose id could not be parsed still reports `created`.
    - `error` object: Why this item did not complete. Present when `status` is `failed`, meaning the record was not written, and when `status` is `uncertain`, meaning it is unknown whether it was. Also present, more rarely, alongside `created`/`updated`: the record WAS written and post-write processing failed, so the item reports the error beside its success status. Per-item failures are independent — one item's rejection says nothing about the others. The exception is a failure of the request itself, such as a timeout: where a connector writes the whole batch in one call, every item shares that outcome and carries the same error.
      - `status_code` number: HTTP status code
      - `error` string: Contains an explanation of the status_code as defined in HTTP/1.1 standard (RFC 7231)
      - `type_name` string: The type of error returned
      - `message` string: A human-readable message providing more details about the error.
      - `detail` union: Contains parameter or domain specific information related to the error and why it occurred.
        - string
        - object
      - `ref` string: Link to documentation of error type
    - `warnings` object[]: Non-fatal problems with THIS item — a currency the connector could not match, for example, which is sent without it. Present per item rather than per request, because a caller with twenty records needs to know which one was affected. A warning does not imply the record was written: some are raised while preparing the item, before the write is attempted, so they can appear on a `failed` item too. Read `status` for what happened.
      - `type` string
      - `operation` string
      - `message` string
      - `status_code` number
      - `error` string
  - `_raw` Raw, nullable: Raw response from the integration when raw=true query param is provided

## Other responses

- `400`: Bad Request
- `401`: Unauthorized
- `402`: Payment Required
- `404`: The specified resource was not found
- `422`: Unprocessable
- `default`: Unexpected error

## Changes

- **2026-10-01** `4d9d527eb7ce`: 1 info
  - endpoint added

[Operation history](https://skmtc.dev/apideck-libraries/apis/accounting-api/changes/accounting/tracking-categories/batch/post.md)

---

[API](https://skmtc.dev/apideck-libraries/apis/accounting-api.md) · [All operations](https://skmtc.dev/apideck-libraries/apis/accounting-api/llms.txt) · [OpenAPI document](https://skmtc.dev/apideck-libraries/apis/accounting-api/revisions/d02fc67b26ca?raw)