### WhisperKit Local Model Setup (M-series Macs)

Source: https://context7.com/context7/macwhisper_helpscoutdocs/llms.txt

This guide explains how to install and use WhisperKit models for local transcription on M-series Macs. It details the installation process through the Model Picker, the initial load time, and the benefits of using WhisperKit, including improved accuracy and performance.

```bash
# Requirements: M-series Mac only (M1, M2, M3, M4)
# Not supported: Intel-based Macs

# Installation:
# Model Picker (top right) → Manage Models
# Filter: "WhisperKit"
# Download: Turbo model (recommended)

# First-time load:
# Initial preparation: 2-5 minutes (large models)
# System optimization: One-time process
# Subsequent loads: Instant

# Benefits:
# - Better accuracy on long files
# - Handles silence and music better
# - Faster performance vs older engine
# - Speaker recognition support (green badge)
```

--------------------------------

### Setup Automatic Meeting Recording in MacWhisper

Source: https://context7.com/context7/macwhisper_helpscoutdocs/llms.txt

Guides users on how to enable and configure automatic meeting recording within MacWhisper. This includes granting necessary permissions and selecting applications to monitor. Requires MacWhisper Pro license.

```bash
# Setup meeting recording
# Settings → Record Meetings → Enable "Automatic Meeting Notifications"
# Grant: Screen Recording + Microphone permissions

# Supported applications
# Arc, Brave, Chrome, Chime, Discord, Edge, Firefox, Safari
# Skype, Slack, Teams, Telegram, Webex, WhatsApp, Zoom
# Note: FaceTime not supported (macOS privacy restrictions)

# Configure observed apps
# Settings → Record Meetings → Dropdown menu
# Select/deselect apps to monitor

# Recording workflow
# 1. Start meeting in supported app
# 2. "Meeting Detected" notification appears (10s timeout)
# 3. Click "Record" in notification
# 4. Recording visible in main window
# 5. End meeting → Click stop (bottom right) or in notification
# 6. Recording appears in history sidebar
# 7. Select recording → Auto-transcribe

# Configure speaker names
# Settings → Record Meetings
# "Your Name": Your display name
# "Meeting Participants": Default label for other attendees

# Note: Beta feature - use manual recording for critical meetings
# Available only in macwhisper.com version (not App Store)
```

--------------------------------

### MacWhisper Global Transcription Overlay Setup

Source: https://context7.com/context7/macwhisper_helpscoutdocs/llms.txt

This section outlines the setup for MacWhisper's global transcription overlay. It details how to enable the global mode, configure a system-wide keyboard shortcut, grant microphone permissions, and provides an overview of available configuration options for the floating transcription window.

```bash
# Enable Global mode
# Settings → Global OR Click "Global" on homescreen
# Configure keyboard shortcut (e.g., Cmd+Option+G)
# Grant microphone permissions

# Configuration options
```

--------------------------------

### MacWhisper Dictation with AI Enhancement Setup and Usage

Source: https://context7.com/context7/macwhisper_helpscoutdocs/llms.txt

This section details the setup and usage of MacWhisper's dictation feature with AI enhancement. It includes instructions for basic dictation setup, configuring keyboard shortcuts, granting microphone permissions, and utilizing AI-enhanced dictation with custom prompts for tasks like grammar correction, translation, and email generation. API key setup for OpenAI models is also covered.

```bash
# Setup dictation
# 1. Click "Dictation" on homescreen OR Settings → Dictation
# 2. Configure keyboard shortcut (e.g., Cmd+Shift+D)
# 3. Grant microphone permissions

# Basic usage
# 1. Focus any text field in any application
# 2. Press configured keyboard shortcut
# 3. Speak into microphone
# 4. Text appears in active field

# AI-enhanced dictation (requires Pro license + API key)
# Settings → Dictation → Create custom prompts

# Example prompts:
# "Fix spelling and grammar"
# System prompt: "Clean up the following dictated text, fixing spelling and grammar errors while preserving meaning and tone."

# "Translate to Spanish"
# System prompt: "Translate the following text to Spanish, maintaining formality level."

# "Generate professional email"
# System prompt: "Expand this brief message into a professional customer support email with proper greeting and closing."

# API key setup
# Settings → OpenAI → Add API key
# Requirements: Payment method linked, $1 minimum usage for GPT-4 access
# Models supported: GPT-3.5, GPT-4, GPT-4 Turbo, GPT-4o
```

--------------------------------

### MacWhisper Batch Transcription Setup and Configuration

Source: https://context7.com/context7/macwhisper_helpscoutdocs/llms.txt

This guide explains how to set up and configure batch transcription in MacWhisper. It covers different methods for opening the batch transcription feature, adding multiple export formats with custom settings, viewing duration estimates, and understanding the file naming conventions for exported transcripts.

```bash
# Open Batch Transcription
# Method 1: Click "Batch Transcription" button on homescreen
# Method 2: File → Open File → Select multiple files
# Method 3: Drag multiple files onto MacWhisper window

# Configure export formats
# Click "Add Format" (top right) to add multiple export styles
# Each format can have custom settings
# Example: Add .txt, .srt, .docx formats simultaneously

# View total duration estimate (bottom left)
# Start transcription → All files processed with all selected formats

# Export location: User-defined in settings
# Files named: [original_filename].[format_extension]
```

--------------------------------

### MacWhisper App-Specific Dictation Prompts Configuration

Source: https://context7.com/context7/macwhisper_helpscoutdocs/llms.txt

This guide explains how to configure app-specific AI prompts for dictation in MacWhisper. It covers adding applications to trigger specific prompts automatically, providing examples for Slack, VS Code, and Mail, ensuring tailored AI processing based on the active application.

```bash
# Setup app-specific prompts
# Settings → Dictation → App Specific Prompts

# Add applications
# 1. Ensure target app is running
# 2. Click "+" button
# 3. Select app from list
# 4. Choose prompt from dropdown

# Example configurations:

# Slack (Spanish-speaking channel)
# App: Slack.app
# Prompt: "Translate to Spanish"

# VS Code (coding assistant)
# App: Visual Studio Code.app
# Prompt: "Generate Python code with type hints and docstrings"

# Mail (professional communication)
# App: Mail.app
# Prompt: "Formalize text for business email"

# Usage: Dictate in any app - prompt applied automatically
```

--------------------------------

### MacWhisper License Recovery Steps

Source: https://context7.com/context7/macwhisper_helpscoutdocs/llms.txt

This guide provides steps to recover lost MacWhisper Pro license keys. It instructs users to check their email for purchase confirmations from Gumroad, access their Gumroad library, or contact support with purchase details.

```bash
# Step 1: Check email
# Search inbox for: "You've bought MacWhisper" from Gumroad
# License key included in purchase confirmation

# Step 2: Gumroad Library
# Visit: https://www.gumroad.com/library
# Login with purchase email
# View all purchases and license keys

# Step 3: Contact support
# Email: support@macwhisper.com
# Include: Purchase email address + approximate date
# Team will locate in system and resend
```

--------------------------------

### Cloud Transcription Provider Configuration

Source: https://context7.com/context7/macwhisper_helpscoutdocs/llms.txt

This section details how to set up and manage various cloud transcription providers within MacWhisper. It includes API key configuration, file limits, model information, and pricing for OpenAI, Groq, ElevenLabs, and Deepgram, as well as guidance for custom OpenAI-compatible providers.

```bash
# Configure providers
# Settings → Cloud Transcription → Add API keys

# OpenAI setup
# API Key: sk-proj-xxxxxxxxxxxxx
# File limit: 25MB
# Model: whisper-1
# Pricing: ~$0.006/minute

# Groq setup
# API Key: gsk_xxxxxxxxxxxxx
# File limit: 25MB
# Model: whisper-large-v3
# Pricing: Free tier available

# ElevenLabs setup
# API Key: xxxxxxxxxxxxx
# File limit: 1GB (reliable with good internet)
# Supports: Speaker recognition
# Pricing: Various tiers

# Deepgram setup
# API Key: xxxxxxxxxxxxx
# File limit: 1GB (reliable with good internet)
# Supports: Speaker recognition
# Pricing: Pay-as-you-go

# Custom OpenAI-compatible providers
# Base URL: https://your-provider.com/v1
# API Key: xxxxxxxxxxxxx
# Use for: Llama, Mistral, local inference servers

# Select cloud provider
# Model Picker (top right) → Choose cloud provider
# Transcribe as normal

# Performance considerations:
# 25MB limit: ~40-50 minutes of audio
# 1GB limit: Works best with stable, fast internet
# Slow connection: May fail with large files
# Speaker recognition: ElevenLabs and Deepgram only
```

--------------------------------

### MacWhisper Assistant Features (macOS/iOS)

Source: https://context7.com/context7/macwhisper_helpscoutdocs/llms.txt

This section outlines the features available through the MacWhisper Assistant, including translation, AI chat, summary generation, and cloud transcription. It details availability on iOS and macOS, and explains the pricing model including Pro licenses and optional subscriptions.

```bash
# Assistant features (macOS and iOS)

# Translation
# Translate transcripts to 30+ languages
# Select transcript → Translation → Choose target language
# Available: macOS only

# AI Chat
# Query transcripts conversationally
# "What were the main points discussed?"
# "List all action items mentioned"
# "Who spoke about the budget?"
# Available: iOS and macOS

# Summary generation
# Automatic highlights and key points
# Meeting action items and decisions
# Timestamped important moments
# Section-by-section bullet points
# Available: iOS and macOS

# Assistant Cloud Transcription
# Server-side processing for faster results
# No local computation required
# Ideal for older/slower Macs
# Warning: Audio sent to servers - avoid sensitive content
# Available: iOS and macOS

# Pricing model
# macOS: Pro license required ($39-49 one-time)
# iOS: All features free including local models
# Optional: Assistant subscription for cloud services
# Alternative: Use your own API keys (no subscription needed)

# iOS API key support
# Settings → Cloud Transcription → Add API keys
# Settings → AI Prompts → Add API keys
# Supported: OpenAI, Groq, ElevenLabs, Deepgram
```

--------------------------------

### MacWhisper Transcription Engine Selection and Configuration

Source: https://context7.com/context7/macwhisper_helpscoutdocs/llms.txt

This section details how to select and configure transcription engines within MacWhisper. It covers both local WhisperKit models for M-series Macs and various cloud-based transcription providers like OpenAI, Groq, ElevenLabs, and Deepgram. Supported audio formats and the extensive list of supported languages are also provided.

```bash
# Local models (WhisperKit) - M-series Macs only
# Navigate to: Model Picker (top right) → Manage Models
# Filter by: WhisperKit
# Download: Turbo model (recommended for most Macs)

# After download, first load takes 2-5 minutes for preparation
# Subsequent loads are instant

# Cloud providers - configured in Settings → Cloud Transcription
# OpenAI: 25MB file limit
# Groq: 25MB file limit
# ElevenLabs: 1GB file limit
# Deepgram: 1GB file limit

# Supported audio formats
# mp3, wav, m4a, mp4, mov

# Supported languages (100 total)
# English, Spanish, French, German, Chinese, Japanese, Arabic, etc.
```

--------------------------------

### Configure ChatGPT Integration for Transcripts in MacWhisper

Source: https://context7.com/context7/macwhisper_helpscoutdocs/llms.txt

Provides instructions for integrating ChatGPT with MacWhisper to enhance transcripts with AI-powered summaries, translations, and custom prompts. This involves setting up an OpenAI API key and selecting the desired AI model. Note the new API key requirements from OpenAI.

```bash
# Initial setup
# Settings → OpenAI → Enter API key

# New API key requirements:
# 1. Add payment method to OpenAI account
# 2. Spend $1+ using GPT-4o mini model first
# 3. GPT-4 access unlocked automatically

# Error: "You do not have access to model 'GPT-4o'"
# Solution: Use GPT-4o mini until $1 threshold reached
# Reference: https://help.openai.com/en/articles/7102672

# Select AI model
# ChatGPT Model Picker → Choose model
# GPT-3.5: Fastest, lowest cost
# GPT-4: Higher quality, slower
# GPT-4 Turbo: Balanced performance
# GPT-4o: Latest, most capable

# Using prompts
# 1. Open transcript
# 2. Prompts section → Select pre-defined prompt OR
# 3. Click "Create New Prompt"
#    - Enter prompt description
#    - Enter prompt text
#    - Save

# Example custom prompts:
```

--------------------------------

### Configure Watch Folders for Automatic Transcription in MacWhisper

Source: https://context7.com/context7/macwhisper_helpscoutdocs/llms.txt

Explains how to set up and manage watch folders for automated audio file transcription. This feature allows MacWhisper to automatically transcribe files placed in specified directories. Requires MacWhisper Pro license.

```bash
# Setup watch folders
# Settings → Watch Folders → "Add Watch Folder"
# Select directory to monitor

# Configure auto-transcription
# Toggle: "Auto-Transcribe Watch Folder"
# Enabled: Files transcribed automatically
# Disabled: Notification sent, manual transcription required

# Grant notification permissions when prompted
# System Settings → Notifications → MacWhisper → Enable

# Add export formats
# Click "Add Export Format"
# Select format type (TXT, SRT, DOCX, etc.)
# Expand format → Configure settings
# Add multiple formats (all applied to each file)

# Usage
# 1. Copy/move audio file to watched folder
# 2. MacWhisper detects new file
# 3. Transcription starts automatically (if enabled)
# 4. Exports saved in configured formats
# 5. Output location: User-defined in settings

# Manage folders
# Trash icon: Stop watching folder
# Folder icon: Open in Finder

# Requirements: MacWhisper Pro license
# Status: Beta feature
```

--------------------------------

### Troubleshooting Empty Transcription Errors

Source: https://context7.com/context7/macwhisper_helpscoutdocs/llms.txt

This section addresses the 'Transcription finished without any text' error in MacWhisper. It suggests toggling the 'Downmix to Mono' setting and provides instructions for contacting support with relevant files if the issue persists.

```bash
# Error cause: Audio too short or decode issue

# Solution:
# Settings → Advanced → "Downmix to Mono"
# Toggle ON if currently OFF
# Toggle OFF if currently ON

# Test transcription again

# Still failing:
# Email: support@macwhisper.com
# Include: Audio file (if possible) and model used
```

--------------------------------

### Enable and Manage Automatic Speaker Recognition in MacWhisper

Source: https://context7.com/context7/macwhisper_helpscoutdocs/llms.txt

Details the process of enabling and managing automatic speaker recognition for multi-person audio recordings. This includes selecting compatible models, renaming speakers, merging duplicates, and reassigning segments. Supports local WhisperKit models or cloud providers like ElevenLabs and Deepgram.

```bash
# Requirements
# Local: WhisperKit models (M-series Mac)
# Cloud: ElevenLabs or Deepgram providers
# Not supported: Dictation, "Transcribe Podcast" feature

# Enable speaker recognition
# Select compatible model from Model Picker
# Look for green "Speaker Recognition" badge

# Post-transcription management

# Rename speakers
# Sidebar → Click speaker name → Type new name → Press Enter/unfocus

# Merge duplicate speakers
# Right-click speaker (in transcript or sidebar)
# Select "Merge Speaker To..." → Choose target speaker

# Autofill from previous transcripts
# Type speaker name → Autocomplete suggestions appear
# Press Cmd+Return to select

# Change speaker colors
# Sidebar → Click color palette icon → Select color

# Reassign segments with keyboard shortcuts
# Hover over segment
# Press number key (1, 2, 3, etc.) for corresponding speaker
# Segment reassigned instantly

# Example workflow:
# 1. Transcribe interview with WhisperKit model
# 2. Default labels: "Speaker 1", "Speaker 2", "Speaker 3"
# 3. Rename: "Speaker 1" → "John Smith"
# 4. Notice Speaker 2 and 3 are same person
# 5. Right-click "Speaker 3" → Merge to "Speaker 2"
# 6. Rename "Speaker 2" → "Jane Doe"
# 7. Fix misattributed segments with number keys
# 8. Export with speaker labels intact
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.