whispercpp-kit 0.1.2

A toolkit for whisper.cpp with audio processing and model management

WhisperCPP Kit

A wrapper arround whisper.cpp with additional helper features like model management capabilities.

While whispercpp_kit will automatically build and set up whisper.cpp for you, it requires certain system-level dependencies (git, cmake, ffmpeg) to be pre-installed on your system. If any dependency is missing, the library will provide a clear error message indicating what needs to be installed.

System requirements

Before installing whispercpp_kit, ensure you have the following dependencies installed:

• git
• cmake
• ffmpeg

Installing dependencies

Ubuntu/Debian:

sudo apt update
sudo apt install git cmake ffmpeg

MacOS:

brew install git cmake ffmpeg

Installation

pip install whispercpp_kit

Quickstart

from whispercpp_kit import WhisperCPP

whisper = WhisperCPP(model_name="tiny.en")
# whisper = Whisper(model_path="/custom/pathggml-large-v3-turbo-q5_0.bin")

output = whisper.transcribe("input.mp3")
print(output)

Docker example

See examples/docker/README.md for more details. You can run the container in two ways:

git clone https://github.com/s-emanuilov/whispercpp_kit
cd whispercpp_kit/examples/docker

docker build -t whispercpp_kit .

# Run with default model (base.en)
docker run -v $(pwd):/app/audio whispercpp_kit your_audio.mp3

# Using specific model (works with both mounting options)
docker run -v $(pwd):/app/audio whispercpp_kit your_audio.mp3 tiny.en