TIRITIX
SPEECH · AI · DICTATION · WINDOWS

XoSkryb

AI-powered dictation that types for you — in any application, in any language. Speak naturally, XoSkryb listens, transcribes, and injects the text exactly where your cursor is. Powered by Whisper. Runs entirely on your machine.

Open source
View on GitHub
CAPABILITIES
CORE

Always-On Listening

Runs silently in the background. Detects speech via RMS amplitude monitoring and triggers transcription automatically on natural pauses — no button to press.

INTEGRATION

Types Into Any App

Transcribed text is injected character-by-character into whichever window has focus using native Windows SendInput API. Works with any application — editors, browsers, chat tools.

AI

Powered by Whisper

Uses OpenAI's Whisper model for high-accuracy transcription. Runs locally on your machine — audio never leaves your device.

PERFORMANCE

GPU Acceleration

CUDA support for faster-than-real-time transcription on compatible NVIDIA GPUs. Falls back gracefully to CPU.

LANGUAGE

Multi-Language

Language is configurable. Whisper's multilingual capability means you can dictate in French, English, Spanish, German, and many more without switching tools.

UTILITY

File Transcription

Beyond live dictation, XoSkryb includes a standalone utility to transcribe existing audio files — useful for meeting recordings, voice memos, or interviews.

BUILT WITH
Python
OpenAI Whisper
CUDA / GPU
Windows SendInput
← Back to Apps