Skip to content
Stack-Junkie

EasySpeechToText Privacy Policy

Last updated: March 21, 2026

EasySpeechToText is a desktop speech-to-text application developed by Christopher Kvamme and distributed through the Microsoft Store. This policy explains how the application handles your data.

Core Privacy Principle

EasySpeechToText processes all audio entirely on your device. No audio, transcriptions, or personal data are sent to any server, cloud service, or third party. The application works completely offline after the initial model download.

What the Application Accesses

Microphone

The application accesses your microphone to capture speech for transcription. Microphone audio is processed in real-time by the on-device speech recognition model and is never recorded to disk, transmitted over the network, or retained after transcription is complete.

Keyboard Input (Global Hotkey)

The application registers a global hotkey (default: F9) so you can start and stop recording from any application. This uses the standard Windows RegisterHotKey API. The application does not monitor, log, or capture any other keyboard input.

Clipboard

When auto-paste is enabled, the application briefly uses the system clipboard to insert transcribed text at your cursor position. Your previous clipboard contents are saved and restored immediately after insertion. The application does not read, log, or retain your clipboard contents.

Data Stored on Your Device

  • Settings (model preference, hotkey, audio device, theme) are stored in your user profile directory (%APPDATA%\EasySpeechToText\settings.json).
  • Transcription history (your last 100 transcriptions) is stored locally in %APPDATA%\EasySpeechToText\history.json. You can clear this at any time from within the application.
  • Speech recognition models are downloaded from Hugging Face on first use and cached locally. These are open-source AI models, not personal data.

All stored data remains on your device. None of it is transmitted anywhere.

Data We Collect

None. EasySpeechToText does not collect analytics, telemetry, crash reports, usage data, or any other information. There is no account system, no login, and no server communication (other than the initial one-time model download from Hugging Face).

Third-Party Services

  • Hugging Face hosts the open-source speech recognition models that are downloaded on first use. This download is handled by the huggingface_hublibrary. Hugging Face may log download requests according to their own privacy policy. No personal data is sent with these requests.
  • Microsoft Store handles application distribution, updates, and purchase transactions. Microsoft's handling of Store data is governed by Microsoft's Privacy Statement.

Children's Privacy

EasySpeechToText is not directed at children under 13. The application does not collect any personal information from any user, including children.

Changes to This Policy

This policy may be updated when new features are added. The "last updated" date at the top reflects the most recent revision.

Contact

For privacy-related questions, use the Stack-Junkie contact form.