OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Audio & Voice AI

Whisper Diarization

Whisper Diarization is an open-source Python speech-processing pipeline that transcribes audio and assigns speaker labels to transcript words and sentences. It combines Whisper-based speech recognition with vocal extraction, voice activity detection, speaker embeddings and timestamp alignment.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Whisper Diarization — OpenViST project profile in Audio & Voice AI.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when you have an audio file and want a transcript linked to speaker identities, particularly when you can run Python dependencies and review the output for errors.

Who it is for

  • People processing recordings into speaker-attributed transcripts.
  • Researchers or developers who want a Python-based transcription and diarization pipeline they can run on their own system.

Documented uses

  • Transcribe an audio recording and identify who spoke each sentence.
  • Create speaker-attributed transcripts from conversations or other multi-speaker recordings.
  • Use vocal extraction before speaker embedding as part of a diarization workflow.
Before you choose

Requirements & limits

The public GitHub repository identifies the project as BSD-2-Clause licensed. Its README documents a Python installation and command-line workflow.

  • Python 3.10 or later is specified; Python 3.9 requires manual dependency installation.
  • FFmpeg and Cython are prerequisites for installing the requirements.
  • The parallel NeMo-and-Whisper script is experimental and requires at least 10 GB of VRAM.
  • Overlapping speakers are not addressed, and the README warns that errors may occur.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install and run diarization on an audio file

Install the documented prerequisites and run the standard diarization script on a recording.

Before you start

  • Python 3.10 or later; Python 3.9 requires manually installing requirements one by one.
  • FFmpeg and Cython.
  • The repository files, including constraints.txt and requirements.txt, and an audio file to process.

Steps

  1. Install Cython and FFmpeg using a method listed for your operating system in the README.

  2. Install project requirements with the command below.

    Documented command
    pip install -c constraints.txt -r requirements.txt

  3. Run the command below, replacing the placeholder with the audio file name.

    Documented command
    python diarize.py -a AUDIO_FILE_NAME

  4. If needed, consult the listed options to select a model, language, device or batch size, or disable source separation.

Expected outcome

The pipeline processes the supplied audio for transcription and speaker diarization, associating speakers with transcript segments.

Guide limitations

  • Overlapping speakers are not addressed by the pipeline and may contribute to errors.
  • The README notes that errors can occur.
  • Python 3.9 requires manual installation of requirements.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
14 October 2026
Original website title
MahmoudAshraf97/whisper-diarization · Official homepage title · retained in its source language
How we review sourcesSuggest a correction