Introduction

Welcome to the official documentation for Silkwave Voice!

Silkwave Voice is a lightweight, privacy-focused macOS application for audio recording, transcription, and AI-powered summarization and chat. It uses Apple's on-device speech-to-text models for transcription and Apple Intelligence for summaries and follow-up chat.

Core Capabilities

Versatile Audio Recording

Capture audio from any source to suit your needs:

• Microphone: Record external sound (voice notes, in-person meetings).
• System Audio: Record internal sound directly from apps (browser tabs, video streams) - no third-party audio drivers needed.
• Hybrid Mode: Capture both microphone and system audio simultaneously - perfect for recording interviews or Zoom calls.

On-Device Transcription

Convert speech to text in near-realtime using Apple's on-device speech-to-text models (macOS 26+). Because the process is completely local, it is free, fast, and secure. Supports 10 languages with downloadable language packs.

Import & Export

Import audio files you already have - MP3, M4A, WAV, FLAC, and more - and transcribe them on-device, several at a time, choosing the transcription language for each file. Export any recording, recorded or imported, as a single .zip containing the audio, transcript, and summary. See Importing & Exporting.

AI Summarization

Automatically generate structured summaries of your recordings using Apple Intelligence and the ChatGPT extension via macOS Shortcuts. Summaries include key topics, action items, and an overall theme. See AI Summarization for setup.

AI Chat

Ask follow-up questions about any recording in a dedicated side panel - pull out quotes, clarify decisions, or draft follow-up emails. Conversation history is saved per recording. Powered by ChatGPT through Apple Intelligence; only transcript text is sent. See AI Chat for details.

Menu Bar App

Access recording controls directly from the macOS menu bar without opening the main window. Start, pause, and resume recordings, switch audio sources, and change languages - all from a compact popover.

Global Search

Search across all your past recordings and transcripts instantly. Find specific keywords or phrases across your entire library to quickly locate what was said.

Privacy First

Silkwave Voice is built with privacy as a priority.

• On-Device Transcription: All audio recording and transcription happens locally. Your audio files never leave your device.
• No Cloud Required for Transcription: No internet connection needed for transcription. No data sent to external servers.
• AI Features: When using AI Summarization or AI Chat, the transcript text is shared with ChatGPT through Apple Intelligence. Your original audio is never uploaded - only the text transcript is sent, and only when you choose to use these features.

Requirements

• macOS 26+
• Microphone Permission: Required to capture external audio.
• System Audio Permission: Required to capture internal audio from applications, browsers, and video calls.
• Shortcuts Events Permission: Required for AI Summarization and AI Chat. See Permissions & Setup for details.