Documentation

Documentation

Quickstart

Get up and running with Voiceavel in under a minute.

Hold, Speak, Release

  1. Hold your hotkey (Right Command ⌘ by default)
  2. Speak naturally in any supported language
  3. Release the key — your words appear wherever your cursor is

That's it. No clicking, no switching apps, no copy-pasting.

Choosing a Provider

Voiceavel supports five transcription providers:

Uses Groq's Whisper Large v3 on Voiceavel's servers. No API key needed — just select it and go. All AI features (Smart Cleanup, Command Mode, etc.) work automatically.

Local AI (Free)

Uses faster-whisper running entirely on your Mac. No internet required, completely private. Great for English and major languages.

OpenAI (Pro)

Uses GPT-4o Transcribe for the best accuracy across all languages. Requires your own OpenAI API key. Costs ~$0.006/minute.

AssemblyAI Real-time (Pro)

Text appears as you speak — no waiting after releasing the hotkey. Uses AssemblyAI's streaming API via WebSocket (~300ms latency). Requires your own AssemblyAI API key ($50 free credit on signup). Supports English, Spanish, French, German, Italian, and Portuguese for streaming.

Groq BYOK (Pro)

Uses Whisper Large v3 on Groq's ultra-fast hardware with your own API key. Near-instant results with a generous free tier.

AI Features

Voiceavel includes intelligent post-processing that makes your transcriptions cleaner and more useful:

  • Smart Cleanup — Removes filler words (um, uh, you know) and handles mid-sentence corrections automatically
  • Command Mode — Select text, hold your hotkey, and speak an instruction like "make this professional" to transform it
  • Context-Aware Formatting — Adjusts tone based on the app you're in (casual in Slack, professional in Mail, technical in VS Code)
  • Coding Mode — Preserves technical terms and formats identifiers in camelCase when you're in a code editor

AI features work automatically with Cloud (Voiceavel) provider. For OpenAI or Groq BYOK, they use your API key. Free users get basic filler removal without any API key. Toggle features individually in Settings → AI Features.

Click the Voiceavel menu bar icon to find:

  • Dictation History — Your last 30 transcriptions, click any entry to copy it
  • Cloud Usage — Minutes used today and this month (Cloud provider only)
  • Copy Last — Quickly copy your most recent transcription

Changing Settings

Click the Voiceavel menu bar icon to access settings:

  • Provider — Switch between Cloud, AssemblyAI, Local, OpenAI, and Groq
  • Language — Set a specific language or use auto-detect
  • Keyboard Shortcut — Choose from Right Command, Left Command, Right Option, Left Option, or Fn key
  • API Keys — Configure your OpenAI, Groq, or AssemblyAI API keys
  • AI Features — Toggle Smart Cleanup, Context-Aware Formatting, and Coding Mode
  • Audio Feedback — Enable start/stop beep sounds (off by default)

Tips

  • Set your language explicitly for better accuracy, especially if you always speak the same language. Auto-detect works well but manual is more reliable.
  • First transcription warm-up — The very first transcription after launching the app may be slightly less accurate (Local provider). This is normal as the AI model loads into memory.
  • Voiceavel updates automatically — The app checks for new versions every 5 minutes and will notify you when an update is available.