GitHub Repo · Meetings & Notes

Whisper

Open-source speech recognition you can run locally.

What Whisper does

Whisper is OpenAI's open-source speech-to-text model, which transcribes and translates audio in many languages and can run on your own hardware for private transcription.

Best for

Private transcription
Multilingual audio
Developers building voice features

How small businesses use Whisper

Transcribe recorded calls privately
Caption videos
Convert voice memos to notes

Quick facts
Type	GitHub Repo
Category	Meetings & Notes
Pricing	Free and open source
Repository	github.com/openai/whisper
License	MIT
Setup level	Developer
Main tradeoff	It runs from code or the command line, so it needs a developer, and long files are slow without a good graphics card.
GitHub stars	111k (as of 2026-10-21)
Added	Oct 16, 2026
Last reviewed	Oct 16, 2026

Not sure Whisper is right for you?

Garner Digital Consulting helps small businesses pick and set up the right AI stack.

Get help choosing →

Whisper: common questions

Who should use Whisper?

Whisper suits private transcription, multilingual audio and developers building voice features.

When should you choose something else instead of Whisper?

Whisper runs from code or the command line, so it needs a developer, and long files are slow without a good graphics card. If that matters for your business, compare other options such as Fathom, Otter.ai and Loom.

Is Whisper free?

Whisper pricing: Free and open source. Pricing changes often, so confirm current plans on the official repository or website before you buy.

Can a small business self-host Whisper?

Yes. Whisper is open source on GitHub, and its setup level is developer. You will need a computer or server to run it and someone to keep it updated and backed up, or you can use a hosted option if one is offered.