All Articles

deepseek r1 + kokoroTTS + OpenWebUI

架构

deepseek-r1-kokoroTTS-OpenWebUI-199df59e1e4e8045bbdbe7f06565a9be-1.jpg

Install TTS

# install UV
curl -LsSf 
https://astral.sh/uv/install.sh
 | sh
source $HOME/.local/bin/env
uv

# install espeak-ng (ompact open source software text-to-speech synthesizer)
apt-get install espeak-ng

# install ffmpeg 
# FFmpeg is a free and open-source software project consisting of a suite of libraries and programs for handling video, audio, and other multimedia files and streams
apt install ffmpeg

# download source code
git clone 
https://github.com/remsky/Kokoro-FastAPI.git

cd Kokoro-FastAPI

# may need install
pip install loguru

# download model
python docker/scripts/download_model.py --output api/src/models/v1_0

# start app
./start-gpu.sh

在Open WebUI里面指定TTS

注意: 必须是English UI

URL http://localhost:8880/v1
API not-needed
Model kokoro
Voice zm_yunxi

可以使用的Voice

https://huggingface.co/hexgrad/Kokoro-82M/blob/main/VOICES.md