ianblenke / whisper-playground

Build real time speech2text web apps using OpenAI's Whisper https://openai.com/blog/whisper/

Whisper Playground

Instantly build speech2text apps in 99 languages using OpenAI's Whisper

Live demo https://whisperui.monsterapi.ai

Whisper.Playground.mp4

Contribution ideas

• Stream audio using web sockets over the current approach of incrementally sending audio chunks
• Implement diarization (speaker identification) using pyannote-audio (example)

Setup

1. Whisper requires the command-line tool ffmpeg and portaudio to be installed on your system, which is available from most package managers:

# on Ubuntu or Debian
sudo apt update && sudo apt install ffmpeg
sudo apt install portaudio19-dev

# on Arch Linux
sudo pacman -S ffmpeg
sudo pacman -S portaudio

# on MacOS using Homebrew (https://brew.sh/)
brew install ffmpeg
brew install portaudio

# on Windows using Chocolatey (https://chocolatey.org/)
choco install ffmpeg

# on Windows using Scoop (https://scoop.sh/)
scoop install ffmpeg

2. Clone or fork this repository
3. Install the backend and frontend environmet sh install_playground.sh
4. Run the backend cd backend && source venv/bin/activate && flask run --port 8000
5. In a different terminal, run the React frontend cd interface && yarn start

License

This repository and the code and model weights of Whisper are released under the MIT License.