Download draft transcription

get

https://api.verbit.co/api/job/draft

Download the latest available transcription before job completion.

Use Cases

Access to machine-generated captions while waiting for professional review
Early content preview

Availability Check

Polling:

Check draft_ready field in /job/info response
true indicates draft is available
false means no draft yet
Updates throughout processing

Webhook:

Receive a webhook notification with draft_ready: true

Output Options

Same formats as /job/get_caption
All language options supported
Timecode features available
Multiple file formats

Important Notes

Content may be incomplete
Subject to further editing
Not final transcription
Updates periodically

Recent Requests

Time	Status	User Agent
Retrieving recent requests…

Loading…

Query Params

integer

required

API version. Current version is 4. Older versions are supported but deprecated.

job_id

integer

required

Job ID (as received after calling POST /job/new)

caption_format

string

enum

The format of transcription/caption file to return.

Supported formats: srt, web_vtt, sami, txt, docx, pdf.

Default: srt or txt (depending on whether caption segmentation was included in job's
profile).

Allowed:

language

string

The language of transcription/caption file to return.

Format should be RFC 5646 Language code, like 'fr-FR'.

Default value will be job original language code, like 'es-US'.

timecode_interval

integer

Include timecode every X seconds.
Timecode format is [HH:MM:SS] (example: [00:12:59]).
The value should be a positive integer.
It represents the time interval (in seconds) between inserted timecodes.
For example, with timecode_interval=120, timecode will be inserted every 2 minutes.
If the value is not provided, or not numeric, timecodes will not be inserted.
Supported only for text formats: txt, docx, pdf.

timecode_on_speaker_change

boolean

Defaults to false

Include timecode before every speaker change.
Timecode format is [HH:MM:SS] (example: [00:12:59]).
Available only for text formats: txt, docx, pdf and for jobs with speaker names enabled.
Default: false.

inaudible_mode

string

enum

Define the way how should render inaudible.
Could be one from: timestamp, no_timestamp, remove.
If it not defined - use the default behavior.
NOTE: this param works only for txt and docx content types.
For other types - it will not have any impact

Allowed:

Headers

string

enum

Defaults to text/plain

Generated from available response content types

Allowed:

Responses

200success

401unauthorized

404not found