Voice & AudioIntermediate
62.6kPython+63/daypushed 1d ago
Use it for: Clone a voice and train a text-to-speech model from about one minute of voice data using a web UI.
Quick start
- Clone the repo and follow the README install section (Python 3.10-3.12)
- Alternatively, try the Colab training notebook or Docker image
- Launch the WebUI
- Provide a short voice sample and train or run text-to-speech
56.2kPython+307/daypushed 0d ago
Use it for: Clone or design voices, dub videos, dictate, transcribe and create audiobooks locally in 646 languages, as a free ElevenLabs alternative.
Quick start
- On macOS or Linux run: curl -fsSL https://voicestudio.sh/install | sh
- On Windows run in PowerShell: irm https://voicestudio.sh/install | iex
- Launch the desktop app
- Choose a workspace such as voice cloning, dubbing or voice design
Voice & AudioIntermediate
46.1kPython+20/daypushed 784d ago
Use it for: Generate speech from text with pretrained models in many languages, or train and fine-tune your own text-to-speech models.
Quick start
- Install the TTS package from PyPI (pip install TTS).
- Pick a pretrained model, such as ⓍTTS, from the README or docs.
- Generate speech from text with the library or its command-line tools.
- To customize a voice, follow the example fine-tuning recipes in the repo.
Voice & AudioIntermediate
39.9kPython+46/daypushed 182d ago
Use it for: Turn text into natural-sounding conversational speech in English or Chinese, suited to LLM assistants and dialogue.
Quick start
- Clone the repo or install the ChatTTS package from PyPI.
- Download the model weights from Hugging Face.
- Run the example scripts or the Colab notebook to synthesize dialogue speech.
- Try community end-user projects listed in Awesome-ChatTTS for ready-made apps.
Voice & AudioIntermediate
38.5kPython+99/daypushed 2d ago
Use it for: Generate multilingual speech in 30 languages, design new voices from text descriptions, and clone voices from short reference clips.
Quick start
- Clone the repo and follow the README install section.
- Download the VoxCPM2 model weights as described in the README.
- Enter text, optionally with a voice description or reference audio, to generate 48kHz speech.
- Provide reference audio plus its transcript if you want the most faithful cloning.
Voice & AudioIntermediate
37.8kPython+36/daypushed 538d ago
Use it for: Clone a voice's tone color from a short reference clip and generate speech in multiple languages with controllable style.
Quick start
- Clone the repo and follow the README install section.
- Download the OpenVoice V2 checkpoints as directed in the README.
- Supply a reference voice clip and the text you want spoken.
- Adjust style settings such as emotion, accent, rhythm, and pauses.
Voice & AudioIntermediate
36.9kPython+20/daypushed 220d ago
Use it for: Clone a voice from a few seconds of audio and generate arbitrary speech, with strong Mandarin support.
Quick start
- Install Python 3.7+, PyTorch, and ffmpeg.
- Run pip install -r requirements.txt to install the remaining dependencies.
- Prepare the pretrained models as described in the README.
- Launch the toolbox or web server to clone a voice and synthesize speech.
Voice & AudioIntermediate
24.4kPython+40/daypushed 10d ago
Use it for: Clone a voice from a single audio clip and synthesize speech in Chinese, English, Japanese, Spanish or Arabic with emotion and speed control.
Quick start
- Clone the repo and follow the README install section.
- Download the IndexTTS-2.5 model from HuggingFace or ModelScope.
- Provide one reference audio clip and your text.
- Tune emotion, speaking speed, or pronunciation controls as needed.
23.9kPython+29/daypushed 137d ago
Use it for: Synthesize multilingual speech with zero-shot voice cloning, covering inference, training, and deployment, including many Chinese dialects.
Quick start
- Clone the repo and follow the README install section.
- Download a pretrained model such as Fun-CosyVoice 3.0 from ModelScope or HuggingFace.
- Run inference with a text prompt and optional reference voice.
- Use the training and deployment docs to fine-tune or serve the model.
Voice & AudioIntermediate
19.4kPython+36/daypushed 324d ago
Use it for: Generate realistic English dialogue from a transcript in one pass, including nonverbal sounds like laughter and coughing.
Quick start
- Clone the repo and follow the README install section, or use Hugging Face Transformers.
- Download the Dia-1.6B weights from Hugging Face.
- Write a transcript of moderate length, using non-verbal tags sparingly.
- Optionally condition on reference audio for emotion and tone, then generate the audio.
