OpenAI Whisper: Accurate and Free Transcription for Audio and Video Files

A surprisingly good tool for transcribing audio and video files

Mad Machine
ILLUMINATION’S MIRROR

--

Photo by Kane Reinholdtsen on Unsplash

If you need a free and accurate software to transcribe audio or video files, you’re in luck! OpenAI offers Whisper, a tool that transcribes with ease and accuracy.

Whisper supports multiple languages. You can see a list of languages on this link.

This article will provide detailed steps for installing and using this tool. It will cover the installation process on Mac, which should be very similar on Linux. Since the tool is open source, it should also work on Windows, but the procedure will be different. I can provide the procedure for Linux and Windows if requested.

Where to get it?

Whisper is available at openai/whisper: Robust Speech Recognition via Large-Scale Weak Supervision (github.com).

Requirements

Whisper requires these tools:

  • Python 3
  • ffmpeg

Installation

Python3 and ffmpeg are required, as mentioned in the requirements. The easiest way to install them is by using Homebrew. To install, just run these commands:

--

--

Mad Machine
ILLUMINATION’S MIRROR

I’m a DevOps engineer, but I love to tinker with all kinds of technologies.