← Discover
Discover · GitHub · VibeVoice

Analyzing long audio recordings with artificial intelligence

Published by Microsoft, VibeVoice was developed as an open source voice AI framework. With its Python-based structure, the system allows users to train their own sound models and integrate them into their applications.

Updates

  • August 2, 2026: Stars 48,569 → 51,860.

What you get

  • Converts up to 60 minutes of audio recording to text at a time.
  • It provides speaker ID, timestamp and content details in a structured way.
  • Provides user-defined keyword support for custom terms and names.

Installation

Install from GitHub
git clone https://github.com/microsoft/VibeVoice.git
cd VibeVoice
pip install -e .

Running it

Gradio demo
python demo/vibevoice_asr_gradio_demo.py --model_path microsoft/VibeVoice-ASR --share
Transcription from file
python demo/vibevoice_asr_inference_from_file.py --model_path microsoft/VibeVoice-ASR --audio_files [ses-dosyasi]

If you don't write code

🤖 Paste this into your AI agent (Claude Code · Codex · Antigravity)

I want to analyze the 60-minute audio recording I have using the VibeVoice model. I need to retrieve who the speakers are, when they spoke, and the content they said as a structured text file. I also want to add custom keywords so that the model recognizes technical terms more accurately, how can I structure this process?

Related dictionary terms

Artificial Intelligence
Who it is forIt is suitable for users who want to quickly and structuredly convert long-term audio recordings, meeting summaries or podcast content into text.
LicenseMIT

Links

TreScout did not build this tool · we found it in GitHub trends and wrote it up. This page describes the repository as of 2026-06-07: The star count and our text belong to that day, the repository may have changed since. Check the repository link for the current state. This page was machine-translated from the Turkish original · the Turkish version prevails.