LOW COST 
AUTOMATED TRANSCRIPTION

HOW DOES IT WORK?

Easy:

1. Upload an audio file

2. Click "Transcribe" - your file will be inserted into the processing queue

3. Wait until processing completes

4. View or download the transcript

You can upload new files without waiting for the processing of the first file to finish.

Processing each file takes about 1/6 of the audio duration. For example, a file with 1 hour of audio will take about 10 minutes of processing time.

WHAT AUDIO FORMATS DO YOU SUPPORT?

Currently we support:

• WAV

  • PCM, uLaw, or aLaw format​

  • Sample rates between 8kHz and ​48kHz

  • 8-bit and 16-bit sample size

• MP3

• FLAC

Audio in Stereo will we down-mixed to Mono. Maximum size of audio file is 128MB.

If your file format is not listed above, you can submit a request via our "Feature Request Poll"

HOW MUCH DOES IT COST?

First 10 hours of transcription is free each month.

Transcription cost after the 10 free hours is only $0.01 (1 cent) per minute of audio.

HOW DO I PAY?

If you stay within the 10 hours you do not need to pay anything, you even do not have to provide a Payment Method.

​

Beyond that, to be able to submit new audio for transcription you have to maintain a positive Account Balance. You can refill the balance in increments of $5, $10, and $25 using the Payment Method you provided. 

​

To provide the details of the Payment Method you want to use, just click on the gear icon ⚙ under the Account Balance.

WHAT PAYMENT METHODS DO YOU ACCEPT?

Our Billing System accepts MasterCard, Visa, and Amex.

IS TIMING INFORMATION INCLUDED IN TRANSCRIPTION?

It is possible to download four different types of files: plain TXT, SRT, VTT, and JSON.

All of them contain timing information

​

The JSON transcription output contains both the start time and duration for each word. Additionally, it contains confidence value for each word.

HOW LONG ARE THE RESULTS STORED?

The results of transcription are stored for 7 days.

After that time, transcription results expire and are no longer accessible.

Please make sure to download transcripts you want to keep before they expire.

​

If the ability to store the results for longer is important to you, you can submit a request via our "Feature Request Poll"

ANY LIMITATIONS IN THE CURRENT VERSION?

Speech.Works is a product under development and we are adding new features often.

Currently, the main limitations are:

• Only English language is supported

• Does not distinguish between multiple speakers

• Stereo channels are not handled individually but mixed into mono

• There is a built-in review tool, but no built-in editor.

DO YOU PROVIDE A WEB API?

Yes, if you would like to use a web API please submit a ticket via the portal (click ? in the header) and we will upgrade your account to allow access to Web APIs.