Offline dictation
Speak naturally and turn recordings into usable text with on-device speech detection and recognition.
Private AI Voice Writing for Mac
Speak once. Write Anywhere.
Vvicat Write turns your voice into polished text, translations, answers, and reusable writing memory on your Mac, without giving up privacy.
- Local
- ASR + LLM
- Private
- Audio stays on Mac
- Fast
- Shortcut first
Offline engine ready
Voice, rewrite, translate
A Focused Writing Space
For the moment before the sentence disappears.
Capture a thought, clean it up, translate it, ask questions about selected text, or insert it directly into the app you are already using. Vvicat Write keeps the workflow close to the keyboard and on your Mac.
Feature Overview
Voice writing for the full loop.
Write Anywhere
Insert finalized text into the current Mac app with keyboard-first controls designed for repeated use.
Ask Anything
Use your voice with selected text for questions, summaries, and drafting help without leaving context.
Translate and Polish
Turn rough speech into cleaner writing, or translate ideas while preserving intent and tone.
Personal Dictionary and Memory
Keep names, terms, and recurring writing preferences close to the local writing engine.
Capability Layer
Local AI components, tuned for writing.
Vvicat Write combines local voice segmentation, speech recognition, Qwen ASR, and llama.cpp-powered text refinement in a single Mac workflow. Not another chatbot window. Better writing where you already work.
01
Voice activity detection
Finds stable speech segments before recognition.
02
Local ASR
Transcribes speech on the device using bundled models.
03
Writing refinement
Polishes, translates, and answers with local model support.
04
Mac insertion
Moves final text back into your active writing surface.
Product Preview
Quiet surfaces. Fast intent.
Real product screens for setup, recording, review, and model control.
Privacy Model
Built for local-first processing.
Vvicat Write requests microphone access for dictation, Accessibility access for text insertion, and Input Monitoring for shortcuts. The product is designed around local audio processing and local model execution.
Microphone
Used to capture audio for on-device dictation.
Accessibility
Used to insert finalized text into the focused app.
Input Monitoring
Used for configurable local writing shortcuts.
Local Models
Offline speech and writing components remain under your control.
Ready for Mac
Turn speech into finished writing.
Download links, support URLs, and the Privacy Policy can be connected to your App Store Connect metadata when the listing is ready.