What it does
VibeVoice is a set of open AI models for working with voice. Some models do speech recognition: they listen to audio and write down what was said. Others do text-to-speech: they read text aloud in a natural-sounding voice.
The speech recognition model can handle up to an hour of audio in one go. It notes who is speaking, when they speak and what they say. You can give it special words, like names or technical terms, to help it get them right. The README says it supports more than 50 languages. There is also a streaming version that writes text while the audio is still coming in, and a smaller version that can run on an ordinary processor without a graphics card.
A lightweight real-time text-to-speech model starts speaking quickly and can read text as it arrives. A larger long-form speech model was also released, which could create conversations up to 90 minutes long with up to four speakers. The makers say they removed that model's code after finding it was used in ways that didn't match its intended purpose.
Who it's for
Researchers and developers working on transcription or speech generation, and curious users who want to try the online playground or Colab notebooks.
Can I use it without coding?
With some setup. You can try some models in an online playground or a Colab notebook, but running them yourself takes technical setup. New to this? Start here.
Can I run it?
- Cost
- FreeNothing to pay or sign up for.
- Runs on
- Web browser
- Privacy
- Your choice: run it on your device or use an online service
Try it
- Open the VibeVoice-ASR online playground linked in the README to try speech-to-text.
- Try the real-time text-to-speech model in the Colab notebook linked in the README.
- Download the model files from the project's Hugging Face page to run them yourself.
Using Claude Code, Cursor or another AI coding assistant? Paste the prompt and it will walk you through the setup.New to this?
Before you install
- Only download from the official GitHub page. Copies with similar names exist for many popular projects.
- Plain-AyI is an independent guide and doesn't host or review the software it lists. Please check each project's official page before installing.
- License (MIT): Very permissive: free to use, change and share, including for business.
- More on staying safe
Where people are talking about it
- VibeVoice-ASR-Streaming ReportResearch paper · Sept 2026
- VibeVoice-ASR-BitNet ReportResearch paper · Jul 2026
- VibeVoice: Open-source frontier voice AIHacker News · Apr 2026
- VibeVoice-ASR Technical ReportResearch paper · Jan 2026
- Microsoft open weights VibeVoice TTS supports 90 minutes speech, 4 speakersHacker News · Aug 2025
- VibeVoice-TTS Technical ReportResearch paper · Aug 2025