Whisper
by OpenAI
Open-source speech recognition model for transcription and translation.
About
Whisper is OpenAI's open-source automatic speech recognition model. It transcribes audio in many languages and can translate speech into English. You run it yourself, and many other apps build on it.
More pictures
The picture the site shows when it’s shared
Taken from github.com. The pictures belong to the site’s owner and are shown to identify the site.
Useful features
- Multilingual transcription
- Speech translation into English
- Several model sizes to trade speed for accuracy
- Runs locally on your hardware
Good for
- Transcribing interviews and podcasts
- Subtitles
- Building speech features into apps
Pricing
Free. Free and open source. You run it on your own hardware.
Why it’s an editor’s pick
Voice and audio: Open-source speech recognition you can run on your own hardware.
More in Voice and audio
-
#1
Realistic AI voices for speech, voice cloning, dubbing, sound effects and music.
-
#2
Edit video and podcasts by editing the transcript, with AI cleanup and clipping.
-
#3
Otter.ai (moving down)
Meeting assistant that records, transcribes and summarizes calls in real time.
-
#5
Browser-based tools that make voice recordings sound cleaner and more studio-like.
-
#6
Text to speech app that reads documents, web pages and books aloud.