Whisper by Mr D

A native macOS voice-to-text application with global hotkey capture, AI-powered transcription, and smart text optimization. Press Option+X anywhere, speak, and your words appear as text — instantly copied to your clipboard.

Overview

Whisper by Mr D is a lightweight, always-available transcription tool that lives in your menu bar. It captures audio from your microphone, transcribes it using Groq (free) or OpenAI, and automatically copies the result to your clipboard. One hotkey, zero friction.

Why Groq?

Groq offers a free tier for their Whisper API — making this app completely free to use for most users. Groq's LPU inference engine delivers transcription results significantly faster than traditional GPU-based solutions, often completing in under a second.

Provider	Model	Speed	Cost
Groq (default)	whisper-large-v3-turbo	Ultra-fast	Free tier available
OpenAI	whisper-1	Standard	Pay-per-use

Features

Core

• Global Hotkey — Option+X (fully customizable) to start/stop recording from anywhere
• Instant Transcription — Speech-to-text via Groq or OpenAI Whisper API
• Auto-Clipboard — Transcribed text is automatically copied and optionally pasted into the active app
• Live Preview — Real-time transcription feedback every 5 seconds while recording
• Menu Bar Integration — Persistent status icon with quick actions

AI-Powered

• Text Optimization — One-click grammar and style improvement using Llama 3.3 (via Groq)
• Smart Organization — AI-powered folder categorization for your transcriptions
• PRD Generator — Turn voice memos into structured Product Requirement Documents
• Chat Assistant — Conversational AI with voice input (speech-to-speech via ElevenLabs TTS)

Interface

• 3 Window Modes — Compact (mic only), Medium (list), Full (all features)
• 4 Themes — Cyberpunk, Dark, Light, Slate
• Folder System — Organize transcriptions with drag & drop
• Waveform Display — Real-time audio visualization while recording

Productivity

• Transcription History — Searchable archive with favorites, tags, and summaries
• Command Launcher — Execute shell scripts directly from the app
• Script Dashboard — Manage and run custom automation tiles

Requirements

• macOS 13 (Ventura) or later
• Microphone access
• API key: Groq (recommended, free) or OpenAI

Installation

Build from Source

# Clone the repository
git clone https://github.com/yourusername/whisper-by-mrd.git
cd whisper-by-mrd

# Build with Swift Package Manager
cd "whisper APP"
swift build -c release

# Run the app
.build/release/WhisperRecorder

Using the Build Script

cd "whisper APP"
chmod +x BUILD_APP.command
./BUILD_APP.command

This will compile the app, package it into Whisper.app, and launch it automatically.

API Key Setup

Option 1: Groq (Recommended — Free)

1. Go to console.groq.com/keys
2. Create a free account
3. Generate an API key
4. In the app: Settings > select Groq > paste your key > Save

Option 2: OpenAI

1. Go to platform.openai.com/api-keys
2. Generate an API key (requires billing setup)
3. In the app: Settings > select OpenAI > paste your key > Save

API keys are stored securely in macOS UserDefaults (protected by the system keychain). They never leave your machine except to authenticate with the chosen API.

Architecture

whisper APP/
├── Package.swift                    # SPM configuration
├── Sources/WhisperRecorder/
│   ├── WhisperApp.swift             # App entry point & menu bar
│   ├── Models/
│   │   ├── TranscriptionModel.swift # Core data model
│   │   ├── Theme.swift              # 4 theme definitions
│   │   ├── TreeNode.swift           # Folder hierarchy
│   │   └── UIStateModel.swift       # Window mode state
│   ├── Services/
│   │   ├── GroqAPIService.swift     # Groq Whisper transcription
│   │   ├── OpenAIService.swift      # OpenAI Whisper fallback
│   │   ├── TranscriptionManager.swift  # Core orchestrator
│   │   ├── AudioRecorderManager.swift  # AVAudioRecorder wrapper
│   │   ├── TextOptimizationService.swift # AI text enhancement
│   │   ├── GroqChatService.swift    # Conversational AI
│   │   ├── DatabaseManager.swift    # SQLite persistence
│   │   └── AIOrganizer.swift        # Smart categorization
│   └── Views/
│       ├── ContentView.swift        # Mode router (compact/medium/full)
│       ├── MicrophoneMinimalView.swift # Compact mode
│       ├── SettingsView.swift       # API keys & hotkey config
│       ├── ChatView.swift           # AI chat interface
│       └── ...                      # 15+ view components
└── Resources/
    └── AppIcon.iconset/             # App icons

Dependencies

Package	Version	Purpose
HotKey	0.2.1	Global keyboard shortcuts
SQLite.swift	0.15.4	Chat history persistence

Usage

1. Launch the app — it appears in your menu bar as a waveform icon
2. Press Option+X (or your custom hotkey) to start recording
3. Speak — you'll see a live waveform and partial transcription
4. Press Option+X again to stop — the transcription is copied to your clipboard
5. Paste anywhere with Cmd+V

Window Modes

Mode	Size	Use Case
Compact	200x220	Just the mic button — minimal footprint
Medium	400x700	Transcription list with search
Full	1100x800	All features: sidebar, chat, dashboard

Customizing the Hotkey

Open Settings (Cmd+,) > Hotkey section > Click the key field and press your desired combination > Save.

Privacy

• Audio is recorded locally and sent only to the API you choose (Groq or OpenAI)
• Transcriptions are stored locally on your Mac
• API keys are stored in macOS UserDefaults (system-protected)
• No analytics, no telemetry, no cloud sync

Tech Stack

• Language: Swift 5.9
• UI: SwiftUI
• Audio: AVFoundation
• AI: Groq API (Whisper, Llama 3.3), OpenAI API
• Database: SQLite (via SQLite.swift)
• Build: Swift Package Manager

License

MIT License. See LICENSE for details.

Author

Built by Mr D — part of the Infinity Cloud ecosystem.