# CamoVoice - Private Speech-to-Text Software

## Overview

CamoVoice is a desktop speech-to-text application designed with one core principle: **your voice stays on your device**. Unlike cloud-based transcription services, CamoVoice processes everything locally using bundled AI models — no internet connection required, no data leaves your computer, and absolutely zero telemetry.

**Website**: https://camovoice.com/
**Company**: CamoText LLC (https://camotext.ai)
**Contact**: contact@camotext.ai
**Version**: 0.1.1
**Platforms**: Windows, macOS
**Additional Resources**: 
- User Guide: https://camovoice.com/userguide.html
- Why Private Speech-to-Text?: https://camovoice.com/why-private-speech-to-text.html

## Core Value Proposition

CamoVoice is a fully private, offline speech-to-text application that enables users to:
- Dictate AI prompts faster than typing
- Transcribe sensitive audio with complete privacy
- Work entirely offline without internet connectivity
- Avoid text entry logs, third-party telemetry, and cloud data exposure
- Process audio locally using bundled AI models

## Privacy & Architecture

### Privacy Features
- **100% Offline**: All transcription happens on your machine using faster-whisper, which runs entirely locally
- **Zero Telemetry**: CamoVoice collects nothing. No usage analytics, no crash reports, no audio samples
- **No Cloud Dependencies**: Once installed, the app works without any network connection
- **Local Settings**: Preferences stored in a simple `settings.json` file on your device
- **No Accounts**: No sign-ups, no authentication required
- **No Data Transmission**: Audio never leaves your device, transcriptions never sent to servers

### Technical Architecture
- **Speech Recognition**: Uses faster-whisper (CTranslate2-based implementation), 4-6x faster than original Whisper
- **Two Model Options**: 
  - Fast mode: base.en model (~140MB) with greedy decoding
  - Thinking mode: distil-large-v3 model (~1.4GB) with beam search and voice activity detection
- **Models Bundled**: All models included with application, no download required after installation
- **Processing**: CPU-based with int8 quantization for efficiency
- **Audio Capture**: Records at 16kHz mono (Whisper's native sample rate) using SoundDevice
- **Text-to-Speech**: Uses pyttsx3, interfaces with system's speech synthesis (SAPI5 on Windows, NSSpeechSynthesizer on macOS)
- **UI Framework**: CustomTkinter with dark theme and orange accent colors

## Key Features

### Recording & Input
- **Multiple Recording Methods**: Click microphone button or hold spacebar (push-to-talk style)
- **Recording Duration Timer**: Appears after 1 minute, shows elapsed time and mode limit (e.g., `3:45 / 10:00`), color-coded (gray/orange/red) as limit approaches
- **Audio File Support**: Load and transcribe WAV and MP3 files
- **Audio Level Indicator**: Real-time visual feedback showing input audio levels during recording
- **Automatic Processing**: Transcription begins automatically after recording stops
- **File Size Limits**: 60 MB (Fast mode) or 30 MB (Thinking mode) to prevent freezing

### Transcription Modes
- **Fast Mode**: 
  - Speed: ★★★★★
  - Accuracy: ★★★☆☆
  - Best for: Quick notes, short recordings, real-time feel
  - Recording limit: 10 minutes
- **Thinking Mode**:
  - Speed: ★★★☆☆
  - Accuracy: ★★★★★
  - Best for: Important transcriptions, longer recordings, complex vocabulary, accents, background noise
  - Recording limit: 5 minutes

### Export & Save Options
- **Export Formats**: TXT, DOCX, PDF
- **Copy Text**: Copies all text to clipboard (button flashes green to confirm)
- **Timestamp Options**:
  - Include date at top of document
  - Include ISO 24-hour format timestamp after each transcription (e.g., `[Recorded: 09-Jan-2026 14:32:15]`)
  - Optional display of timestamps in-app (controlled by settings)
- **Save Options**:
  - Include edits to transcription (when enabled, timestamps are disabled)
  - Manual text editing supported before saving

### Accessibility Features
- **Scalable Text Size**: Slider adjusts font size throughout entire application (text, buttons, labels, status messages)
- **Text-to-Speech Playback**: Reads transcriptions aloud using system-installed voices
- **Keyboard Shortcuts**:
  - Hold spacebar to record
  - Ctrl+S / ⌘S to save
  - Ctrl+Z / ⌘Z to undo clear
  - Escape to close dialogs (Settings, Save As) without saving changes
- **High Contrast Dark Theme**: Designed for readability and accessibility

### User Interface
- **Simple Interface**: Clean, minimalist design focused on ease of use
- **Visual Feedback**: 
  - Pulsing microphone button during recording
  - Recording duration timer (appears after 1 minute, color-coded as limit approaches)
  - Animated "Transcribing..." status with cycling dots
  - Text area border pulses with accent color during processing
  - Audio level meter during recording
  - Copy button flashes green when text is copied
- **Text Editing**: Full editing capabilities in transcription area
- **Undo Clear**: Restore most recently cleared text (button or keyboard shortcut)
- **Window Size**: Minimum 640×480, responsive layout that scales

### Settings & Configuration
- **Input Device Selection**: Choose from available microphones
- **Playback Voice Selection**: Choose from system-installed TTS voices
- **Show Timestamps**: Toggle display of timestamps in transcriptions
- **Keep Window Always on Top**: Toggle to keep CamoVoice window visible above other applications
- **Custom Words**: Add up to 100 custom vocabulary words or phrases to improve recognition of specialized terms, names, jargon, conditions, organizitions, acronyms, or unusual spellings
  - Words are suggestions that increase recognition likelihood (not guaranteed)
  - Case-preserved but duplicate-checked case-insensitively
  - Saved automatically and persist between sessions
  - Accessible via Settings → Custom Words (n)
- **Auto-Save Preferences**: All settings automatically saved and loaded

## Use Cases

### Primary Use Cases
- **AI Prompt Dictation**: Dictate prompts to AI assistants faster than typing
- **Legal Dictation**: Transcribe legal notes, case summaries, client meetings
- **Medical Records**: Transcribe patient notes, medical dictation
- **Personal Journaling**: Voice-to-text journal entries with privacy
- **Accessibility**: Voice input for users with mobility or typing limitations
- **Sensitive Content**: Any transcription where privacy and data security are paramount

### Ideal For
- Users who need offline transcription capabilities
- Professionals handling sensitive information (legal, medical, financial)
- Users concerned about cloud-based service privacy
- People who want to avoid text entry logs and telemetry
- Users who need to work without internet connectivity
- Accessibility-focused users requiring scalable text and TTS features

## Pricing & Versions

### Pro Version
- **Price**: $24.99 USD (one-time payment)
- **License**: Pay once, use forever with free updates for one year
- **Features**:
  - 100% offline, no dependencies, no data leaves device
  - Fast and Thinking bundled models, English-optimized
  - Optional ISO-timestamped transcriptions
  - Save as TXT, DOCX, or PDF
  - Windows and macOS support
  - Full accessibility features
  - Audio file loading (WAV, MP3)
- **Download**: https://camotext.lemonsqueezy.com/checkout/buy/065acea6-45d6-40d8-b163-15176c8b2d5a

### Enterprise Version
- **Pricing**: Custom quote (contact for pricing)
- **Additional Features**:
  - 17+ languages supported (English, Spanish, French, German, Italian, Portuguese, Dutch, Russian, Chinese, Japanese, Korean, Polish, Turkish, Ukrainian, Arabic, Hindi, Vietnamese)
  - Option to auto-detect input language
  - Option to auto-translate to English
  - Additional file formats for input sounds and output files
  - Higher file size limits
  - Bespoke interface and settings customization
  - Enhanced audit trail features and export formats
- **Contact**: contact@camotext.ai

## Technical Specifications

### System Requirements
- **Operating Systems**: Windows, macOS
- **Installation Size**: ~1.2GB (includes bundled models)
- **Internet**: Not required after installation
- **Microphone**: Any system-compatible microphone
- **Text-to-Speech**: Requires system-installed TTS voices (available through OS settings)

### Supported Audio Formats
- **Input**: WAV (any sample rate/bit depth), MP3
- **Processing**: Automatically resampled to 16kHz mono
- **Not Supported**: M4A/AAC (must convert to WAV or MP3)

### Performance Characteristics
- **First Transcription**: May take longer as model initializes
- **Subsequent Transcriptions**: Faster after model is loaded
- **Mode Switching**: Requires loading different model (slight delay)
- **Processing**: CPU-based, optimized with int8 quantization

## Limitations

### Current Limitations
- **Language Support (Pro)**: English only
- **File Size Limits**: 60 MB (Fast mode), 30 MB (Thinking mode)
- **Recording Time Limits**: 10 minutes (Fast mode), 5 minutes (Thinking mode)
- **M4A/AAC Files**: Not supported (must convert to WAV or MP3)
- **Text-to-Speech**: Requires system-installed voices

### Workarounds
- Break up long recordings to avoid time limits
- Convert unsupported audio formats before loading
- Install additional TTS voices through OS settings if needed

## Frequently Asked Questions

### Why use a private speech-to-text app?
Private speech-to-text apps like CamoVoice protect sensitive information by avoiding text entry logs on websites, preventing third-party telemetry, and ensuring complete data containerization. Your voice and transcriptions never leave your device, making it ideal for AI prompt creation, legal dictation, medical records, and personal notes. No internet connection is required.

### How does CamoVoice work?
CamoVoice comes with AI models fully bundled with the application. The initial app load may be slower as the models are loaded into memory, but afterwards they are cached on your device and will work seamlessly. All transcription processing happens locally on your computer using faster-whisper technology, requiring no internet connection after installation.

### How fast is transcription?
Fast mode provides near-instant results using a compact model, perfect for quick notes and short recordings. Thinking mode uses a larger model for better accuracy with complex vocabulary, accents, and background noise. The first transcription may take slightly longer as the model initializes, but subsequent transcriptions are faster.

### Do I need to configure anything?
No configuration is required—just install and start dictating. CamoVoice works out of the box with your system's default microphone. Optional settings include selecting a specific input device, choosing a playback voice, and adjusting text size. All settings are automatically saved.

### What limits are there to recording or loaded audio files?
Recordings have time limits: 10 minutes for Fast mode, 5 minutes for Thinking mode. Audio files have size limits: 60 MB for Fast mode, 30 MB for Thinking mode. Break up recordings or files to avoid limits.

### What formats and timestamp options are available?
CamoVoice supports saving as TXT, DOCX, or PDF. You can include a date header at the top of documents and optional ISO 24-hour format timestamps after each transcription segment. You can also choose to save manual edits, which disables per-transcription timestamps since edited text may no longer match original segments.

## Comparison to Alternatives

### Advantages Over Cloud-Based Services
- Complete privacy: no data transmission
- Works offline: no internet required
- No accounts or sign-ups
- No usage limits or subscription fees (one-time payment)
- No telemetry or analytics
- Suitable for sensitive content

### Advantages Over Other Offline Solutions
- Bundled models: no separate download required
- Two optimized modes: speed vs. accuracy
- Multiple export formats: TXT, DOCX, PDF
- Accessibility features: scalable text, TTS playback
- Simple interface: minimal learning curve
- English-optimized: maximum accuracy for English transcription

## Company Information

**Company Name**: CamoText LLC
**Website**: https://camotext.ai
**Email**: contact@camotext.ai
**Description**: CamoText LLC specializes in privacy-first AI tools and software. We build offline, local-processing applications that protect user data while enabling powerful AI capabilities.
**Founded**: 2025
**Social Media**: 
- LinkedIn: https://www.linkedin.com/company/camotext/
- GitHub: https://github.com/camotext

## Keywords & Search Terms

speech to text, offline transcription, private speech-to-text, private transcription, private transcriber, local speech-to-text software, private AI, voice dictation, AI prompts, accessibility, privacy-focused software, CamoText, CamoVoice, faster-whisper, offline dictation, local transcription, private voice recognition, desktop transcription, speech recognition software, voice-to-text offline, secure transcription, confidential transcription, legal dictation software, medical transcription software, AI prompt dictation, privacy-first transcription

## Related Products & Services

CamoVoice is part of the CamoText product line, which focuses on privacy-first AI tools. The company may offer additional products in the future related to private AI processing and offline machine learning applications.

---

*Last Updated: 2026-01-27*
*For the most current information, visit https://camovoice.com/*