How to Transcribe Text for Free Using AI

Long before AI was used to generate video and code programs, it was used to understand and act on spoken language. AI is the reason you can chat with Alexa, Siri, and Google Assistant and have them do your bidding.

These same algorithms can help you create digital transcriptions of audio files with speech, whether those files are meetings, interviews, lectures, or just voice notes you’ve recorded for yourself. Transcripts give you a written record of what was said and are easily searchable.

Major transcription services such as Rev And Happy writer only offer a limited amount of transcription for free, but you can get the job done without paying anything thanks to Whisper. This is the speech-to-text engine developed by OpenAI (of ChatGPT fame) and there are no restrictions on its use.

You have the option to use a web version of the hosted app at Hugging Facewhich is convenient, but often busy at peak times. You can also choose to install the software locally on Windows. This means faster transcriptions, but you will need a decent PC to handle the AI ​​processing requirements.

Whispering on the web

Whisper on the web is fast and easy to use. Screenshot: Whisper

Go to Whispering on cuddly faceand you can have audio transcribed in your browser for free — you don’t even need to sign up for an account. You have the option to upload an audio file from your computer, or you can record speech directly into the app if you have a microphone connected. Be aware that your audio may be used to further train future AI models — as is often the case, OpenAI and Hugging Face’s privacy policies are unclear on this.

Uploading and processing an audio file:

  • Open the Sound file tab.
  • To elect Click to upload.
  • Select an audio file.
  • Check the Transcribe box.
  • Click Submit.

After a few moments (or longer) you will see the text output on the right side of the screen. Processing times will vary depending on the length of your audio file and how busy the Hugging Face servers are. Since this is a free service open to everyone, it is also very popular, so you may have to wait a long time for files to come through the queue.

Within the interface, you’ll find a few handy tools. For example, click the little pen icon just above the audio playback bar and you can trim the beginning and end of the clip. Handy if you want to cut out silences or unimportant parts of the audio.

You can also switch to the Microphone tab to record some audio directly into the Whisper interface, or switch to the YouTube tab and get transcripts of any video. Just paste the video URL and you're good to go. It's worth noting that YouTube already automatically adds transcripts to some videos, which you can find in the comments sections.

Whispering on Windows

screenshot of code
Whisper uses a simple command prompt interface. Screenshot: Whisper

If you experience a lot of latency in the Whisper web app, or if you just want to keep your transcription processing local and private, you can set up the AI ​​model on a Windows machine. You'll need a CUDA-compatible graphics card (see here for a list), with at least 4GB of VRAM to handle the processing. If you have a fairly recent Nvidia card installed, you probably qualify.

This process is much more involved and doesn't give you much in the way of a user interface, so it's not for everyone. It does have the advantages we've already mentioned, and you won't be stuck in a queue waiting for your files to be processed. It's also a fun little project if you like tinkering with code and programs.

Assuming your computer meets the requirements, you will need to install a number of components on your computer: Python for the coding (make sure that Add python.exe to PATH will be checked during installation), PyTorch for the machine learning libraries, Chocolatey for managing software packages, and FFmpeg for audio processing. They all come with installation instructions on the relevant websites, if you need them.

You are then ready to install Whisper yourself: Search for “cmd” in the Start menu and open Command Promptthen type “pip install -U openai-whisper” and press EnterOnce the installation is complete, you can transcribe files as follows:

  • Open the folder containing your audio files in File Explorer.
  • Click on the address bar at the top, type “cmd” and press Enter.
  • Type “whisper” followed by a space and the name of your audio file.
  • Hit Enter and the processing begins.

The text will be displayed on the screen and saved as a series of text files in the same folder as the audio. If you want to convert multiple files at once, simply list them all after the “whisper” command, separating each file with a space.

Even if you’re not familiar with Python or the command line, you shouldn’t have too much trouble getting everything up and running. There are plenty of tutorials online that can help you out if you need help. This is one of the best tutorials outside, taking you step-by-step through each stage and explaining some of the advanced features available to you (such as switching to a different AI model).

Related Posts

  • Technology
  • July 12, 2024
  • 1 views
  • 14 minutes Read
Top Benefits of DevOps You Can’t Afford to Ignore

by Bipin Mishra Since its introduction, DevOps methodologies have sparked widespread interest in the IT industry, driven by the significant benefits they bring to project pipelines. This blog delves into…

How Carhartt Reimagined a '50s Work Jacket and Became an Unlikely Fashion Icon

The Carhartt Detroit work jacket is an unlikely darling of the vintage clothing world. It’s been produced in various forms for 70 years, seen in auto shops, on construction sites,…

Leave a Reply

Your email address will not be published. Required fields are marked *

You Missed

Bezoek een landhuis van 60 miljoen dollar in Delray Beach, Florida

  • July 12, 2024
Bezoek een landhuis van 60 miljoen dollar in Delray Beach, Florida

Asian Markets: US CPI, Yen Intervention

  • July 12, 2024
Asian Markets: US CPI, Yen Intervention

AT&T says hackers have gained access to the phone calls and text messages of nearly all of its mobile customers

  • July 12, 2024
AT&T says hackers have gained access to the phone calls and text messages of nearly all of its mobile customers

How to Spend the Best Sunday in LA, According to Lisa Ann Walter

  • July 12, 2024
How to Spend the Best Sunday in LA, According to Lisa Ann Walter

Hydrogen flight looks set for takeoff with new developments

  • July 12, 2024
Hydrogen flight looks set for takeoff with new developments

Top Benefits of DevOps You Can’t Afford to Ignore

  • July 12, 2024
Top Benefits of DevOps You Can’t Afford to Ignore

Wells Fargo WFC Q2 2024 Earnings

  • July 12, 2024
Wells Fargo WFC Q2 2024 Earnings

Elon Musk's X misleads users and violates online content rules, says EU

  • July 12, 2024
Elon Musk's X misleads users and violates online content rules, says EU

'Sing Sing' Avoids Simplified Stories of Hope: NPR

  • July 12, 2024
'Sing Sing' Avoids Simplified Stories of Hope: NPR

What previous U.S. Soccer hires tell us about the plan to replace fired USMNT coach Gregg Berhalter

  • July 12, 2024

American tourist dies after suddenly falling ill on Mount Etna in Sicily, rescuers say

  • July 12, 2024
American tourist dies after suddenly falling ill on Mount Etna in Sicily, rescuers say