Whisper Real Time Transcription

This is a demo of real time speech to text with OpenAI's Whisper model. It works by constantly recording audio in a thread and concatenating the raw bytes over multiple recordings.

Setup

Create and activate a virtual environment, then install dependencies:

python3 -m venv whisper_venv
source whisper_venv/bin/activate
pip install -r requirements.txt

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

# on Ubuntu or Debian
sudo apt update && sudo apt install ffmpeg

# on Arch Linux
sudo pacman -S ffmpeg

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

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

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

For more information on Whisper please see https://github.com/openai/whisper

Running the Application

The web UI serves at http://127.0.0.1:5000 (also http://localhost:5000) by default. There are two ways to run this application:

Web UI Mode

To run the application with a web-based user interface:

python3 web_output.py

Common web UI commands with explicit models and microphone backends:

python3 web_output.py --model tiny --default_microphone pipewire
python3 web_output.py --model base --default_microphone pipewire
python3 web_output.py --model small --default_microphone pipewire
python3 web_output.py --model medium --default_microphone pipewire
python3 web_output.py --model large --default_microphone pipewire
python3 web_output.py --model large-v3-turbo --default_microphone pipewire

Gemini API key variants:

python3 web_output.py --model large-v3-turbo --default_microphone pipewire --gemini_api_key YOUR_GEMINI_API_KEY
python3 web_output.py --model large-v3-turbo --default_microphone pipewire --gemini_api_key "$(cat .env | grep GEMINI_API_KEY | cut -d= -f2)"
python3 web_output.py --model large-v3-turbo --default_microphone pulse --gemini_api_key "$(cat .env | grep GEMINI_API_KEY | cut -d= -f2)"

OpenRouter example:

python web_output.py --model large-v3-turbo --default_microphone pipewire --openrouter_api_key "$(cat .env | grep OPENROUTER_API_KEY | cut -d= -f2)"