Whisper
Open-source speech recognition from OpenAI that transcribes and translates audio.
Open sourceAudio & Voice
Overview
Whisper is a speech-to-text model you can run on your own machine. It handles many languages, including Arabic, and can translate speech into English text.
pip install -U openai-whisper
whisper lecture.mp3 --language ar
Features
- Transcription in many languages
- Speech translation to English
- Runs offline
- Several model sizes for speed or accuracy
Use cases
- Transcribing lectures and meetings
- Creating subtitles
- Making audio archives searchable
Articles about this tool
Free, freemium, paid, open source: what the labels mean
Every tool in the library carries one of four pricing labels. Here is what each one promises — and what it does not.
Running AI on your own computer: a first look
Local models keep your data on your machine and cost nothing per use. Here is what you need and what to expect.