torchaudio
An audio package for PyTorch
Decision gist · record as of 2026-08-14
Yes, if you are building audio or speech ML models with PyTorch and need GPU acceleration and autograd support. The permissive BSD license and active maintenance are favorable. Be aware that it is in maintenance mode (feature removals since 2.8), so verify that remaining APIs match your needs. Install friction is moderate due to compiled wheels, but pre-built distributions are available for common platforms and Python versions.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires PyTorch to be installed separately; compiled wheels require compatible Python version (3.10–3.14) and platform (macOS, Linux, Windows).
- Medium install friction due to compiled wheels for multiple Python versions and platforms (3.10–3.14, macOS ARM64/x86_64, Linux aarch64/x86_64, Windows).
- Active maintenance with recent releases; in maintenance phase since 2.8 with feature removals to reduce redundancy and scope.
License · maintenance · safety
permissive license (permissive) — Permissive license (BSD) allows commercial and private use. Pre-trained models may have separate licenses (e.g., SquimSubjective under CC-BY-NC 4.0); users are responsible for verifying dataset and model permissions.
last release 2026-03-23 (144 days) · last repo commit 2026-08-14 · 2,924 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 12,339,584 downloads/mo, #1,325 on PyPI
Alternatives
Verify before relying
pip install torchaudio
import torchaudio
waveform, sample_rate = torchaudio.load('audio.wav')
spectrogram = torchaudio.transforms.Spectrogram()(waveform)- Whether PyTorch is automatically installed as a dependency or must be installed separately.
- Specific performance characteristics or benchmarks for GPU acceleration on different hardware.
- Current scope and API stability after transition to maintenance phase in version 2.8.
What it is and what it does
torchaudio is a PyTorch extension for audio and speech processing designed specifically for machine learning workflows. It provides GPU-accelerated audio transforms (Spectrogram, MelSpectrogram, MFCC, MuLawEncoding/Decoding, Resample), dataloaders for common audio datasets, forced alignment, and compliance interfaces that align with other libraries like Kaldi. All computations use PyTorch operations, making it feel like a natural extension of the PyTorch ecosystem with full autograd support for trainable features.
The library transitioned into maintenance phase starting with version 2.8, removing redundant features to reduce scope and focus on its core strength: processing audio data for ML. It is not a general signal-processing library but rather a specialized tool for building audio-based machine learning models. It supports modern Python versions (3.10–3.14) across macOS, Linux, and Windows, with compiled wheels for efficient installation.
Use it for
- Build end-to-end speech recognition or audio classification models with GPU acceleration and automatic differentiation.
- Load and preprocess standard audio datasets (with built-in dataloaders) for training neural networks.
- Apply common audio transforms (spectrograms, mel-frequency cepstral coefficients, resampling) as differentiable layers in PyTorch models.
- Align audio with transcriptions using forced alignment for speech processing pipelines.
- Ensure compatibility with Kaldi-based audio processing through compliance interfaces.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you are building audio or speech ML models with PyTorch and need GPU acceleration and autograd support.
The permissive BSD license and active maintenance are favorable. Be aware that it is in maintenance mode (feature removals since 2.8), so verify that remaining APIs match your needs. Install friction is moderate due to compiled wheels, but pre-built distributions are available for common platforms and Python versions.
Install
torchaudio on PyPI
Before you install
Medium install friction due to compiled wheels for multiple Python versions and platforms (3.10–3.14, macOS ARM64/x86_64, Linux aarch64/x86_64, Windows). Active maintenance with recent releases; in maintenance phase since 2.8 with feature removals to reduce redundancy and scope.
Requires PyTorch to be installed separately; compiled wheels require compatible Python version (3.10–3.14) and platform (macOS, Linux, Windows).
License in practice
Permissive license (BSD) allows commercial and private use. Pre-trained models may have separate licenses (e.g., SquimSubjective under CC-BY-NC 4.0); users are responsible for verifying dataset and model permissions.
Quickstart
pip install torchaudio
import torchaudio
waveform, sample_rate = torchaudio.load('audio.wav')
spectrogram = torchaudio.transforms.Spectrogram()(waveform)Verify before relying
- Whether PyTorch is automatically installed as a dependency or must be installed separately.
- Specific performance characteristics or benchmarks for GPU acceleration on different hardware.
- Current scope and API stability after transition to maintenance phase in version 2.8.
Package facts
| License | permissive license permissive |
| Python support | Not specified |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | None |
| Maintenance | Actively maintained 144 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 12,339,584 / month, #1,325 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Environment :: PluginsIntended Audience :: DevelopersIntended Audience :: Science/ResearchLicense :: OSI Approved :: BSD LicenseOperating System :: MacOS :: MacOS XOperating System :: Microsoft :: WindowsOperating System :: POSIXProgramming Language :: C++Programming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.13Programming Language :: Python :: 3.14Programming Language :: Python :: Implementation :: CPythonTopic :: Multimedia :: Sound/AudioTopic :: Scientific/Engineering :: Artificial Intelligence |
Evidence: torchaudio-2.11.0-cp310-cp310-macosx_11_0_arm64.whl; torchaudio-2.11.0-cp310-cp310-manylinux_2_28_aarch64.whl; torchaudio-2.11.0-cp310-cp310-manylinux_2_28_x86_64.whl; torchaudio-2.11.0-cp310-cp310-win_amd64.whl; torchaudio-2.11.0-cp311-cp311-macosx_11_0_arm64.whl; torchaudio-2.11.0-cp311-cp311-manylinux_2_28_aarch64.whl; torchaudio-2.11.0-cp311-cp311-manylinux_2_28_x86_64.whl; torchaudio-2.11.0-cp311-cp311-win_amd64.whl; torchaudio-2.11.0-cp312-cp312-macosx_11_0_arm64.whl; torchaudio-2.11.0-cp312-cp312-manylinux_2_28_aarch64.whl; torchaudio-2.11.0-cp312-cp312-manylinux_2_28_x86_64.whl; torchaudio-2.11.0-cp312-cp312-win_amd64.whl; torchaudio-2.11.0-cp313-cp313-macosx_12_0_arm64.whl; torchaudio-2.11.0-cp313-cp313-manylinux_2_28_aarch64.whl; torchaudio-2.11.0-cp313-cp313-manylinux_2_28_x86_64.whl; torchaudio-2.11.0-cp313-cp313t-macosx_12_0_arm64.whl; torchaudio-2.11.0-cp313-cp313t-manylinux_2_28_aarch64.whl; torchaudio-2.11.0-cp313-cp313t-manylinux_2_28_x86_64.whl; torchaudio-2.11.0-cp313-cp313t-win_amd64.whl; torchaudio-2.11.0-cp313-cp313-win_amd64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “pytorch audio processing”
- torchaudioProvides PyTorch-based audio processing, transforms, and dataloaders…
- torch-audiomentationsProvides PyTorch-native audio data augmentation transforms that run…
- asteroid-filterbanksProvides PyTorch-based filterbank implementations for audio signal…
Give your agent the search over MCP, or paste the wish link into any chat.
More Artificial Intelligence packages
LiteLLM provides a unified Python interface to call 100+ LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, and others) using OpenAI-compatible API format, available as both a Python SDK and a self-hosted AI Gateway proxy server.
Install it if you need to work with multiple LLM providers or want to centralize LLM routing in your organization.
Client library and CLI tool for downloading, uploading, and managing models, datasets, and repositories on the Hugging Face Hub platform.
Install it if you work with Hugging Face Hub models or datasets.
LangChain provides a framework for building agents and LLM-powered applications by composing language models, tools, and memory through a unified API that abstracts over multiple model providers.
hf-xet provides chunk-based deduplication and efficient file transfer for the Hugging Face Hub, enabling faster uploads and downloads of large files with local disk caching.
Tokenizers converts raw text into token sequences for NLP models, with support for training custom vocabularies and using pre-built tokenizers (BPE, WordPiece) optimized for speed via Rust.
Transformers provides a unified framework for loading, fine-tuning, and running state-of-the-art pretrained models across text, vision, audio, video, and multimodal tasks using PyTorch, JAX, or TensorFlow.
Install it if you need to run or train any transformer-based model for NLP, vision, audio, or multimodal tasks.
See also s3tokenizer · torchlibrosa · lhotse · python_speech_features · openunmix · Gammatone · deepfilternet · laion-clap · torchtext · torchcrepe