vosk-api
by alphacep
Vosk is an offline speech recognition toolkit that turns spoken audio into text without sending it to an online service. It works in more than 20 languages and can be used from Python, Java, Node.js, C# and other languages, on devices from a Raspberry Pi to servers.
- Stars
- 15.2k +3 a day
- Licence
- Apache-2.0
- Runs on
- Your computer
- Setup
- Moderate
- Language
- Jupyter Notebook
- Age
- 2617 days
What you can do with it
- Transcribe lectures, interviews and meetings to text on your own machine
- Generate subtitles for videos without using a paid cloud service
- Add voice commands to a chatbot, smart home device or mobile app
Make money with it
A freelancer could offer transcription or subtitling services, using Vosk to produce first-draft text before checking it by hand.
An idea, not financial advice. Check the licence before charging for anything built on it.
Ways a business could use it
- Transcribe customer calls, interviews or meetings in-house, avoiding per-minute fees and keeping recordings private
- Build voice-controlled features into a product or kiosk without ongoing cloud speech costs
Can you use it commercially?
Free to use, change and sell, including in paid products. Keep the copyright notice.
Not legal advice. Read the project's licence file before building a product on it.
Getting it running
You install it as a library for your language (for example with pip for Python), then download a language model; the Vosk website has instructions and examples.
Follow the official instructions in the project README.
Things to check
- Small models are less accurate than the best paid cloud services, especially with accents or noisy audio
- It is a developer toolkit, not a ready-made app, so some coding is needed
Checked 03 Nov 2026. Summary drafted by AI from the project's README.
