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

WTM (Whisper Turbo MLX)

WTM is a lightweight Python library and command-line tool for transcribing audio with Whisper using MLX inference. It accepts audio files, returns transcription text and segments through its Python function, and offers CLI options for multilingual transcription, timestamps, and quick mode.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

WTM (Whisper Turbo MLX) — 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 WTM when you need a small Whisper-based transcription library with both CLI and Python entry points, and its documented MLX-based approach and installation choices suit your environment.

Who it is for

  • Python developers adding local audio transcription to an application
  • People building command-line workflows for audio files
  • Linux users selecting the documented CUDA or CPU-only installation option

Documented uses

  • Convert an audio file into text from a terminal.
  • Call transcription from Python and receive text with segments.
  • Request multilingual transcription or include timestamps.
  • Add audio transcription to an application or batch utility.
Before you choose

Requirements & limits

  • The README documents standard package installation and distinct Linux CUDA and CPU-only extras.
  • FFmpeg is optional; the library falls back to librosa when FFmpeg is not installed.
  • Quick mode is described as faster but choppier.
  • The repository specifies the MIT license.

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 WTM for Linux transcription

Install the package with the documented Linux option for CUDA or CPU-only use.

Before you start

  • A Linux environment
  • Choose either the CUDA or CPU-only installation extra

Steps

  1. For CUDA, install with the command below.

    Documented command
    pip install whisper-turbo-mlx[cuda]

  2. For CPU-only use, install with the command below.

    Documented command
    pip install whisper-turbo-mlx[cpu]

Expected outcome

The WTM package is installed using the selected Linux extra; the README separately documents CLI and Python usage.

Guide limitations

  • These instructions cover installation only, not verification of the environment or transcription performance.
Documentation 1 ↗
Source-based practical guide

Transcribe an audio file from the command line

Use the documented CLI to request a transcription, optionally with timestamps or multilingual handling.

Before you start

  • WTM installed
  • An audio file such as MP3 or WAV

Steps

  1. Run wtm audio.mp3 for a basic transcription.

  2. Add --multilingual for multilingual transcription or --timestamps to include timestamps.

  3. For a WAV file, wtm audio.wav --quick selects quick mode.

Expected outcome

WTM produces a text transcription; the README example shows timestamped text segments when timestamps are requested.

Guide limitations

  • The README says quick mode is faster but produces choppier output.
  • FFmpeg is optional and recommended for faster decoding; librosa is used as a fallback when FFmpeg is absent.
Documentation 1 ↗
Evidence and transparency

Sources & review

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