---
title: "AudioPiper Docs | ModelPiper"
description: "Documentation for AudioPiper. Speech-to-text, text-to-speech, and audio processing pipelines running locally."
canonical: "https://modelpiper.com/docs/audiopiper"
---

# AudioPiper Docs | ModelPiper

> Documentation for AudioPiper. Speech-to-text, text-to-speech, and audio processing pipelines running locally.

# AudioPiper

Offline

Multi-source audio mixer and recorder — mic, system audio, and per-app capture via Core Audio Taps.

## Capabilities

### Multi-Source Mixing

Capture from microphone, system-wide audio, or individual application audio simultaneously.

### Recording

Actor-based streaming WAV writer with constant memory usage. Waveform visualization with 8x gain.

### Per-App Audio

Core Audio Taps for capturing audio from specific applications — isolate any app's output.

## Install & Permissions

AudioPiper is a free download on the Mac App Store and runs on macOS 26 or later. Per-app audio capture uses Core Audio Taps, which macOS gates behind its own permission prompt.

1.  Install from the [Mac App Store](https://apps.apple.com/us/app/audiopiper/id6759811522) and launch it.
2.  Approve Microphone access when prompted, in System Settings under Privacy & Security, then Microphone. This covers microphone sources only.
3.  For system audio or a single app's output, approve Audio Recording as well. macOS asks separately the first time you add a system or per-app source.
4.  Relaunch AudioPiper once after granting access so the new permission takes effect.

Recording writes a streaming WAV at constant memory, so a long session does not grow the app's footprint. If a per-app source records silence, that app was launched before the tap was approved; relaunch it.