Podcast Transcriptions

Podible’s transcription engine converts speech to text accurately, quickly, and affordably.

Unlike other transcription services, it is tuned specifically for podcasts resulting in higher accuracy.

We can provide back podcast transcriptions for the most popular podcasts in bulk, or for any podcast/audio file via our API.

API

Connect to our API to receive on-demand audio transcripts. Results come back rapidly (typically less than ⅓ the audio duration) and reliably. It is also significantly more affordable than the industry standard (speak to us for the most up to date pricing).

Integration

The API URL format is:
https://api.podscriber.ai/transcribe?url={AUDIO_FILE_URL}&clientId={CLIENT_ID}&timestamps=[true|false]

Parameters:

audio_file_url required; wav, mp3, or m4a audio file url, URL encoded
client_idrequired; your unique ID that Podible provides
timestampsoptional; “true” to get each word’s start time

The API responds with an S3 URL, which contain the status of your transcription and soon the transcript itself.

The S3 file is a JSON object of format:
{
clientId: string;
requestId: string;
status: [Running|Error|Done];
errorMessage?: string;
transcript?: string;

url: string;
}


You can continue polling this file, until status changes from Running to Done or Error. If the status is Done, the text for your audio file will be in the transcript field.

If timestamps was set to false, transcript will be the plain text.

If timestamps was true, transcript will be a JSON encoded string. It will be an array of objects of word/startTime pairs, e.g.:

[{ “word”: “hello”, “startTime”: 0.000}, { “word”: “there”, “startTime”: 0.021}]

Custom

Upon request, Podible can also provide the following (still in beta):

  • speaker splitting (diarization)
  • punctuation
  • brand safety rating
  • keyword tagging

Get a demo!

transcriptions@podible.co