Skip to main content
POST
/
api
/
createTranscription
Create a new transcription job
curl --request POST \
  --url https://app.xosum.am/api/createTranscription \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "metadata": {
    "phonenumber": "+37400000000",
    "agent": "King Ruzi"
  },
  "checklistId": "a1B2c3D4e5F6g7H8i9J0"
}
'
{
  "docId": "<string>",
  "uploadUrl": "<string>"
}
Creates a transcription job and returns a docId plus a pre-signed uploadUrl. Upload your .mp3 to that URL via PUT (the URL is valid for 1 hour) to begin processing. The type field controls how the resulting transcript is shaped:
  • single_voice — plain transcript with no speaker labels. Use for monologues, dictations, voicemails, or any single-speaker audio.
  • phone_call — diarized transcript with inline Խոսնակ 1: / Խոսնակ 2: speaker labels, pinned to exactly two speakers. Use for two-party conversations (support, sales, interview calls).
Once processing finishes, the result is available two ways — pick whichever fits your environment:
  • Webhook (push) — Xosum POSTs the transcript to a URL you host. See Receive Transcription.
  • Polling (pull) — call GET /api/getTranscription with the docId until status is transcribed. Recommended when you can’t expose a publicly reachable webhook receiver.

Authorizations

​
Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Body

application/json
​
type
enum<string>
required

The kind of audio being transcribed. This value also controls diarization behavior in the resulting transcript: single_voice returns a plain transcript with no speaker labels — use this for monologues, dictations, voicemails, single-speaker recordings. phone_call returns a diarized transcript with inline speaker labels (Խոսնակ 1: / Խոսնակ 2:) and is pinned to exactly two speakers — use this for two-party conversations such as support or sales calls.

Available options:
single_voice,
phone_call
​
metadata
object

Optional metadata to attach to the transcription (e.g., user IDs, external info).

Example:
{
  "phonenumber": "+37400000000",
  "agent": "King Ruzi"
}
​
checklistId
string

Optional. The ID of a QA Checklist to run an automated analysis on the transcript. You can find the Checklist ID in the "QA Ստուգացուցակներ" (QA Checklists) section of the Xosum.am web app. The results of the analysis will be available in the "History" page of the web app after the transcription is complete.

Example:

"a1B2c3D4e5F6g7H8i9J0"

Response

Successfully created transcription job and received an upload URL.

​
docId
string

Unique ID for the created transcription job.

​
uploadUrl
string<uri>

A pre-signed URL for uploading the .mp3 audio file using a PUT request. This URL is valid for 1 hour.