Chatterbox

Take the words right out of my mouth with English

AI-powered voice tool that helps non-English-speaking parents create an English language environment for their children.

What it does

When your child speaks (in Chinese or English), Chatterbox:

1. Listens and transcribes their speech
2. Translates and responds in natural, child-friendly English
3. Speaks the English response aloud

Quick Start

Prerequisites

• Python 3.10+
• A microphone
• A DeepSeek API key

Install

git clone https://github.com/raphael-alex/chatterbox.git
cd chatterbox
pip install -r requirements.txt

Configure

# Set your API key
export DEEPSEEK_API_KEY="sk-..."

# Or copy and edit config.yaml
cp config.yaml my_config.yaml
# Edit my_config.yaml with your settings

Run

python main.py

Press Ctrl+C to exit.

Configuration

Edit config.yaml to customize:

Setting	Options	Default
ASR engine	whisper-api, whisper-local	whisper-api
LLM engine	openai, deepseek	openai
TTS engine	edge-tts	edge-tts
TTS voice	Edge-TTS voice names	en-US-JennyNeural
VAD silence duration	seconds	1.5

Using local Whisper (offline ASR)

asr:
  engine: whisper-local
  whisper_local:
    model_size: base  # tiny, base, small, medium, large
    device: auto

Architecture

Microphone → VAD → ASR → LLM → TTS → Speaker

All components use an adapter pattern — swap engines via config without code changes.

License

MIT