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

Allosaurus

Allosaurus is a pretrained Python library and command-line tool that recognizes phones in WAV speech audio across more than 2,000 languages, returning a phone sequence rather than a written-word transcript.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Allosaurus — 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 Allosaurus when a workflow needs phone-level recognition from mono-channel WAV audio, particularly for multilingual speech analysis.

Who it is for

  • Speech and phonetics researchers working with audio across multiple languages.
  • Developers building speech-analysis or phonetic transcription workflows.

Documented uses

  • Generate phone sequences from a WAV audio file.
  • Process a directory of audio files and associate each result with its filename.
  • Choose a language-specific phone inventory or use the full inventory.
Before you choose

Requirements & limits

The public repository identifies the project as open source under the GPL-3.0 license. The package can be installed with pip.

  • Audio must be WAV and mono-channel; the sampling rate may vary because Allosaurus resamples audio for its model.
  • The default ipa setting uses the full inventory of around 230 phones; the project says selecting a language inventory can improve recognition accuracy.
  • The interface supports command-line and Python use.

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 Allosaurus and recognize a WAV file

Set up the package and produce a phone sequence from one mono-channel WAV file.

Before you start

  • Python environment with pip
  • Mono-channel WAV audio

Steps

  1. Install the package with the command below.

    Documented command
    pip install allosaurus

  2. Run the command below with the path to the WAV file.

    Documented command
    python -m allosaurus.run -i

Expected outcome

The recognizer prints a phone sequence to standard output.

Guide limitations

  • The input should be WAV and mono-channel.
  • The default language setting uses the full inventory; a language-specific inventory can be selected with the --lang option.
Documentation 1 ↗
Source-based practical guide

Recognize a directory of WAV files

Run phone recognition over a directory of audio files.

Before you start

  • Installed Allosaurus package
  • A directory containing mono-channel WAV files

Steps

  1. Pass the directory path to the command-line interface using -i.

  2. Optionally select a language inventory with --lang or direct output to a file with --output.

Expected outcome

Allosaurus outputs phone sequences alongside filenames, sorted by filename; output goes to the terminal by default or to the specified file.

Guide limitations

  • Audio files should be WAV and mono-channel.
  • The documentation describes language inventories as an option that can improve recognition accuracy.
Documentation 1 ↗
Evidence and transparency

Sources & review

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