Hold-to-speak Transcribe mode
Hold the always-on-top pill, speak, release — local transcription pastes at your cursor in email, Slack, Word, or IDEs. Professional dictation with auto-punctuation.
AI TRANSCRIPTION
Local AI Transcription — Private Speech to Text
VocalFuse is AI transcription software that runs a local transcription model (~~57 MB) on your Windows PC. Hold-to-speak dictation pastes into any app; Pro adds note-taking modes — without cloud latency or privacy tradeoffs.
From $5/mo · Windows 10+ pill widget · Offline after first download · No audio upload
AI transcription features that matter
Note Taking mode (Pro)
Click to start/stop long sessions — meetings, calls, videos. Each transcript batch and the final AI summary can sync to Account → Notes.
On-device model (~~57 MB)
Downloaded once into C:\VocalFuse\models\. Audio is never sent to a cloud speech API — only text may sync on Pro.
Flat monthly pricing
Skip per-minute cloud bills. Basic $5/mo unlimited dictation; Pro $10/mo adds AI summaries, Continuous mode, and optional Gmail delivery.
Three recording modes
| Mode | Plan | What it does |
|---|---|---|
| Transcribe | Basic / Pro | Hold pill → local paste at cursor (works without key) |
| Note Taking | Pro | Click start/stop → AI summary + Account → Notes sync |
| Continuous | Pro | Background 60-second clips → local bullets only |
Local AI transcription vs cloud APIs
Cloud AI transcription charges per minute, stores audio remotely, and adds network latency. VocalFuse: flat fee, audio on your machine, transcription local. Need longer capture? VocalFuse Pro adds Note Taking, Continuous, Account → Notes, and Gmail for $10/mo. See Local AI guide.
- ✓ On-device transcription (~57 MB model)
- ✓ Hold-to-speak into any app
- ✓ Note Taking & Continuous (Pro)
- ✓ Account → Notes sync (Pro)
- ✓ Windows 10+ always-on-top pill
- ✓ No per-minute fees
Explore related AI note taking guides
AI Transcription FAQ
Is VocalFuse AI transcription really local?
Yes. A ~57 MB transcription model downloads on first launch into your PC and runs inference on-device for Transcribe, Note Taking, and Continuous modes. No cloud speech API processes your audio.
Does AI transcription work offline?
After the model downloads, hold-to-speak dictation works offline. You need internet for license checks, first model download, Pro AI summaries, and optional Account → Notes sync.
What is included in Basic vs Pro?
Basic ($5/mo): unlimited Transcribe mode (hold the pill, paste at cursor). Pro ($10/mo): adds Note Taking with AI summaries, Continuous mode, Account → Notes sync, and optional Gmail delivery.
How fast is local AI transcription?
Because inference runs on your CPU (no cloud round-trip), dictation feels responsive once the model is loaded. Faster machines reduce “Working…” time between hold and paste.
Can VocalFuse transcribe pre-recorded files?
VocalFuse transcribes live microphone audio in Transcribe, Note Taking, and Continuous modes. It does not import audio files from disk.