Why Qevoxyl

Simple, installer-grade desktop transcription with local privacy.

Private & Offline

Transcribe locally with Whisper. Audio stays on your machine by default—no cloud uploads.

Installer-Grade UX

Per-user install (no admin), Start Menu & Desktop shortcuts, and a right-click Explorer action.

Fast on CPU, Faster on GPU

Runs well on CPU with fp32; leverages CUDA for fp16 acceleration when available.

Record or Import

One-click recording, or open existing WAV/MP3. Copy text or read it back with TTS.

How it works

From mic to text in a few clicks—no cloud required.

Record or Upload

Capture audio with the built-in recorder, or choose an existing file.

Choose a Model

Pick a Whisper size to balance speed and accuracy.

Transcribe Offline

Qevoxyl runs Whisper on-device using FFmpeg for decoding.

Copy or Speak

Copy cleaned text or let the TTS read it back.

Download

Per-user installer, no admin required. Windows 10/11 (x64).

Qevoxyl Setup (Recommended)

Includes Start Menu/desktop shortcuts and Windows Explorer right-click integration.

Download Installer

System Requirements

Windows 10/11 (x64)
VC++ 2015–2022 runtime
~200MB+ disk (model-dependent)
Internet (first model download)

Integrity

SHA-256 (installer): 7C554B02CBE927C0DB8BAC46CF52B86756195A83C4A720CFC260215D4DD8B479

Verify checksums after download for extra assurance.

What’s inside

Offline Whisper transcription
FFmpeg built-in (fallback)
Right-click “Transcribe with Qevoxyl”
Optional Edge TTS playback
Local model caching

Changelog

v1.3.0 — Quality & stability update

Online TTS toggle (off by default) with saved settings.
Reset/New Session (Ctrl+N), safer async session guards.
FFmpeg fallback, logging, UI hardening, no-UPX packaging.

View full changelog →

Privacy Highlights

Qevoxyl is built for privacy. Here’s the quick version:

✓ Offline by default — all transcription runs on your PC. No cloud uploads.
✓ Your data stays local — audio, text, and models are stored on your device.
✓ No tracking — we don’t collect personal data, telemetry, or usage analytics.
✓ Optional only — internet is needed once for model download, or if you enable TTS.
✓ You’re in control — delete your recordings and transcriptions anytime.

Download Qevoxyl Read Full Privacy Policy

Speech-to-Text Keywords

Offline transcription software, Windows speech recognition, privacy-focused voice transcription, Whisper desktop app, local speech-to-text, no cloud transcription, secure voice recognition, Windows transcription software, offline voice-to-text, private speech recognition

FAQ

Answers to common questions.

Transcription runs on-device and does not need the internet. A connection is only needed the first time you download a Whisper model, and for Edge TTS (if you enable it).

Under your user profile at %LOCALAPPDATA%\Qevoxyl\ (e.g., models, recordings, voice, logs). This avoids admin prompts and keeps data per-user.

Yes. If CUDA is available, Qevoxyl uses fp16 inference for speed. Otherwise, it runs on CPU in fp32 with good compatibility.

Yes. The installer adds a shell context menu for .wav and .mp3. Right-click a file and choose “Transcribe with Qevoxyl.”