Deep Dive

Built for privacy.
Engineered for speed.

Every feature is designed to keep your data on your device while delivering professional-grade transcription.

Privacy architecture

Your audio stays on your device at every stage of processing.

Microphone Local capture
AI Engine On-device models
Text Output Your app
Local DB On-device storage
Core Engine

Advanced AI. 25+ languages (up to 99 with specialty editions). Advanced hallucination removal.

Our transcription pipeline uses a 6-stage process to deliver accurate results: voice activity detection, model inference, hallucination stripping, dictionary injection, text shortcut expansion, and AI grammar polish.

  • Advanced AI engine for speed and accuracy — automatically configured
  • AI grammar polish — on-device AI corrects punctuation, capitalization, and grammar
  • 25+ languages (up to 99 with specialty editions) with automatic detection or manual selection
  • Hallucination stripping removes phantom text from silent segments
  • Filler word removal — automatically strips "uh", "um", "like", "you know" for clean, professional output
  • Custom dictionary injection for specialized vocabulary (medical, legal, technical)
  • Text shortcuts — speak or type trigger abbreviations that expand to full text with dynamic macros

6-Stage Pipeline

VAD Detection
Model Inference
Hallucination Strip
Dictionary Inject
Snippet Expand
AI Polish
Desktop App

Live-typing into any application

VoicePrivate types transcribed text directly into whatever app has focus — no copy-paste needed. Works with every macOS and Windows application.

  • Global hotkey — start transcribing from any app instantly
  • Intelligent silence detection — auto-stops when you pause
  • Transcription history — searchable archive of all your dictation
  • File import — transcribe audio/video files from disk
0:00
Text Shortcuts

Say less, type more

Create trigger abbreviations that expand into full text blocks — activated by voice or keyboard. Just say "slash soap" during dictation and get a complete SOAP note template, or type the trigger directly. Built-in macros insert today's date, time, and clipboard contents automatically.

  • Voice or keyboard triggers — say "slash sig" or type /sig to expand full text blocks instantly
  • Dynamic macros — %date%, %time%, %day%, %clipboard% resolve at expansion time
  • Unlimited snippets — create as many shortcuts as you need, organized by category
  • Import & export — share snippet libraries as JSON, or import from colleagues
  • Built-in text shortcuts — no extra subscription needed

Example Shortcuts

/sig → Best regards,
  John Smith
  Senior Associate
/today%date%
/soap → Subjective:
  Objective:
  Assessment:
  Plan:
/fup → Following up on our
  conversation from %date%.
Compliance

Tamper-evident audit log. Proof audio stayed local.

Zero-trust transcription, verified. Every transcription is recorded in a cryptographic audit trail that proves all audio was processed on-device with zero outbound network calls. Each entry is chained with SHA-256 checksums — if any record is modified, the chain breaks. Export the full log as JSON for compliance reviews.

  • SHA-256 hash chain — each entry's checksum includes the previous entry, making the log tamper-evident
  • Network assertion — every entry records "no_outbound" status proving zero cloud uploads during processing
  • Full metadata — engine used, model path, audio duration, word count, OS version, app version, timestamp
  • JSON and CSV export — one-click export for compliance reviews and internal audits
  • Chain verification — built-in integrity check confirms no entries were altered or deleted
  • Encrypted backup & restore — create encrypted backups of your transcription database and restore on any device, keeping your data safe and portable

Audit Log Entry

timestamp: 2026-03-09T14:32:07Z
event: transcription_complete
engine: on-device-engine
app_version: 1.6.0
os_version: macOS 15.3
source: microphone
duration: 45.2s
words: 312
network: no_outbound
checksum: a1b2c3d4e5f6...
Chain verified — no tampering detected
Advanced

Power features for professionals

Built for the way you actually work — with per-app customization, AI text transformation, multiple export formats, and detailed stats.

  • Per-app modes — different transcription settings per application, automatically applied
  • AI Command Mode — transform text with instructions: formal, casual, bullet points, summarize, fix grammar, or custom prompts
  • Speaker diarization — identify different speakers in multi-person recordings
  • Export formats — TXT, JSON, Markdown, SRT subtitles, and WebVTT
  • Stats dashboard — words transcribed, time saved, WPM, daily streak, top apps, and activity heatmap
  • Quick Recall — instant search of your transcription history with a global hotkey
  • Global text expansion — text shortcuts work system-wide in any application, not just VoicePrivate

AI Command Mode

Input: "we discussed the quarterly revenue targets and agreed to increase marketing spend by fifteen percent while reducing operational overhead in the southeast region..."
Command: Summarize as bullet points
Output:
- Discussed quarterly revenue targets
- Increase marketing spend by 15%
- Reduce operational overhead in SE region

Experience the difference

Start transcribing with complete privacy today.