Overview

Neucodec is an open-source audio-to-audio model designed for high-fidelity neural audio compression and reconstruction. Unlike traditional codecs that rely on rigid mathematical transforms, Neucodec leverages neural architectures to maintain signal integrity while significantly reducing bitrate. For developers, this means an efficient pipeline for transmitting or storing high-quality audio without the typical artifacts of lossy compression. It is particularly useful for integrating into VoIP applications, streaming services, or as a preprocessing layer for downstream speech-to-text and synthesis tasks. Being licensed under Apache-2.0, it offers the flexibility needed for commercial deployment and deep architectural customization.

Highlights

High-fidelity neural audio compression and reconstruction
Apache-2.0 license for flexible commercial integration
Optimized for low-bitrate high-quality audio transmission
Ideal for VoIP and streaming infrastructure pipelines

Usage

Install

# Install Hugging Face transformers
pip install transformers torch

SDK Usage

# Load model with transformers
from transformers import AutoModel, AutoTokenizer

model = AutoModel.from_pretrained("neuphonic/neucodec")
tokenizer = AutoTokenizer.from_pretrained("neuphonic/neucodec")

Hugging Face Download

We recommend downloading the model via the Hugging Face CLI or Hub SDK.

Guidance：Before downloading, install huggingface_hub with:

Guidance

pip install -U huggingface_hub

CLI Download

Download the full repository

huggingface-cli download neuphonic/neucodec

Download a single file to a local folder (e.g. config.json into ./dir)

huggingface-cli download neuphonic/neucodec config.json --local-dir ./dir

See the official docs for more CLI options

SDK Download

# 模型下载
from huggingface_hub import snapshot_download
model_dir = snapshot_download('neuphonic/neucodec')

Git Download

Make sure git-lfs is installed first

Git Download

git lfs install
git clone https://huggingface.co/neuphonic/neucodec

To skip LFS large-file downloads, use:

Skip LFS

GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/neuphonic/neucodec

Model files are hosted on the Hugging Face Hub — download directly via HF CLI / SDK / Git, not through this site.

PyTorch / Transformers Usage

Install Transformers

pip install -U transformers torch

Load the model and run inference

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained('neuphonic/neucodec')
tokenizer = AutoTokenizer.from_pretrained('neuphonic/neucodec')

Model Download

We recommend downloading the model via the ModelScope CLI or SDK.

Guidance：Before downloading, install ModelScope with:

Guidance

pip install modelscope

CLI Download

Download the full repository

modelscope download --model neuphonic/neucodec

Download a single file to a local folder (e.g. README.md into ./dir)

modelscope download --model neuphonic/neucodec README.md --local_dir ./dir

See the docs for more CLI options

SDK Download

# 模型下载
from modelscope import snapshot_download
model_dir = snapshot_download('neuphonic/neucodec')

Git Download

Make sure git-lfs is installed first

Git Download

git lfs install
git clone https://www.modelscope.cn/neuphonic/neucodec.git

To skip LFS large-file downloads, use:

Skip LFS

GIT_LFS_SKIP_SMUDGE=1 git clone https://www.modelscope.cn/neuphonic/neucodec.git

ModelScope 模型页直接下载模型文件；无需将模型文件放在本站服务器。

Notebook Quickstart

Install the ModelScope library

pip install "modelscope[audio,cv,nlp,multi-modal,science]" -f https://modelscope.oss-cn-beijing.aliyuncs.com/releases/repo.html

Load the model and run inference

from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks

p = pipeline('text-generation', 'neuphonic/neucodec')