Skip to content

Installation

Requirements

  • Python 3.10+
  • FFmpeg — for audio processing
  • OpenClaw running locally
  • GPU with CUDA recommended for STT and local TTS, but not required

Clone & Install

git clone https://github.com/ekleziast/kiwi-voice.git
cd kiwi-voice

python -m venv venv
source venv/bin/activate

pip install -r requirements.txt

git clone https://github.com/ekleziast/kiwi-voice.git
cd kiwi-voice

python -m venv venv
venv\Scripts\activate

pip install -r requirements.txt

FFmpeg

sudo apt install ffmpeg

brew install ffmpeg

Download from ffmpeg.org and add to PATH, or set KIWI_FFMPEG_PATH in .env.

Environment File

cp .env.example .env

Edit .env with your API keys. All keys are optional — Kiwi works with free local providers out of the box:

# Optional: ElevenLabs for cloud TTS
KIWI_ELEVENLABS_API_KEY=sk-...

# Optional: Telegram bot for voice security approvals
KIWI_TELEGRAM_BOT_TOKEN=123456:ABC...
KIWI_TELEGRAM_CHAT_ID=123456789

# Optional: RunPod for Qwen3-TTS serverless
RUNPOD_API_KEY=...
RUNPOD_TTS_ENDPOINT_ID=...

Zero-cost setup

Kiwi works fully local and free with Kokoro ONNX (TTS) + Faster Whisper (STT) + text wake word engine. No API keys needed.

Next Steps