offline-stt

🎙️ Production-quality, 100% offline voice typing & AI text refinement for Windows. Hold Ctrl+Space anywhere, speak naturally, and dictate into any app instantly. Powered by .NET 10, Whisper.cpp & LLamaSharp.

Test tiny whisper offline on Android

Lightweight Windows widget that turns microphone speech into clipboard text — Gemini cloud or 6 offline multilingual models, always-on-top, click-through.

The definitive 2026 production guide for mobile Speech-to-Text — 33 sections covering Android, iOS, edge & cloud. whisper.cpp, Moonshine v2, Silero VAD v6, ONNX Runtime. Kotlin · Python · C++.

Voice in + voice out for the Pi coding agent — hold-to-talk STT (Deepgram streaming or offline models) plus TTS (Kitten, Piper, Kokoro, Deepgram Aura). Successor to pi-listen-ng.

Floating Android Overlay for Real-time Audio Summarization. Powered by Vosk (Offline STT) & Google Gemini Flash. Supports Internal Audio recording & YouTube Link summarization. Built with Jetpack Compose & Clean Architecture.

🤖 Assistente de voz privado e modular para Linux com transcrição offline (faster-whisper), servidor WebSockets em Python e aplicativo Android nativo em Jetpack Compose.