SDK

AI Noise Remover SDK

Remove background noise from voice recordings, enhance speech, normalize loudness, convert and measure audio on your own computer or server. Audio never leaves the machine, and processing needs no account or API key. The Python package is loudness-audio-kit.

Install from PyPI View source on GitHub Try it in your browser Installation guide

Speech noise removal

Reduce background noise in voice recordings with a local speech model that runs on the CPU.

Speech enhancement

Noise removal, high-pass filtering, gentle compression and measured loudness normalization in one step.

Loudness normalization

Two-pass normalization to a LUFS target, with true-peak checks and honest warnings when a target is missed.

Conversion

Export WAV, FLAC, MP3 or Opus from common formats, with the sample rate and channels you choose.

Measurement and reports

Loudness, peaks, RMS, silence, channel balance and delivery policies, with offline HTML reports.

Built for services

Configurable limits, bounded concurrent jobs and cancellation that cleans up after itself.

Python

Use Python 3.11 to 3.14 and a local FFmpeg installation. Install the package in a virtual environment, then check that processing works:

python3 -m venv .venv
source .venv/bin/activate
python -m pip install loudness-audio-kit
loudness-audio processing-doctor

Process files from Python:

from loudness_audio_kit import denoise, enhance, normalize, convert

denoise("recording.wav", "clean.wav")
enhance("podcast.mp3", "podcast-clean.wav", target_lufs=-16)
normalize("mix.wav", "mix-normalized.flac", target_lufs=-14, bit_depth=24)
convert("recording.m4a", "recording.opus", bitrate_kbps=128)

The installation guide covers FFmpeg on Windows, macOS and Linux, the first noise removal, the JavaScript build, the local web demo and common problems. Installation guide

Command line

The same operations are available as the loudness-audio command. Results are JSON, and --progress reports progress while a file is processed.

loudness-audio denoise recording.wav --output clean.wav --progress
loudness-audio enhance podcast.mp3 --output enhanced.wav --target-lufs -16
loudness-audio normalize mix.wav --output normalized.flac --bit-depth 24
loudness-audio analyze enhanced.wav > analysis.json

JavaScript and TypeScript

The Node.js edition (Node.js 22 or 24) lives in the same GitHub repository, with TypeScript types. Build it from source:

git clone https://github.com/Loudness-ai/ai-noise-remover-sdk.git
cd ai-noise-remover-sdk/javascript
npm ci --ignore-scripts
npm run build
node dist/cjs/cli.js processing-doctor

Your audio stays local

Processing runs entirely on the computer or server that runs your code. Audio is never uploaded to Loudness.ai, and no Loudness.ai account, API key or server is involved.

Licence

The SDK code is available under the MIT License. The bundled speech model is distributed under its own terms, which the repository lists in DISTRIBUTION-LICENSE and NOTICE.

Prefer the browser?

Use the same kind of processing online, free: