microsoft/VibeVoice
VibeVoice provides open-source frontier voice AI capabilities for developers building audio applications. It is particularly useful for engineers working on text-to-speech and automatic speech recognition systems.
56.2k · Python
You will need to open https://colab.research.google.com/github/microsoft/VibeVoice/blob/main/demo/VibeVoice_colab.ipynb to run the streaming TTS demo without local setup. Adopting this repository means relying on Colab notebooks for quick testing rather than a turnkey local binary, but you get frontier voice AI without paying proprietary API fees.