hns

hns is an open-source command-line speech-to-text tool that prioritizes privacy by processing everything locally on your device. It serves as a direct, open-source alternative to proprietary dictation apps. Your audio never leaves your device - no API keys, no subscriptions, no data collection.

Usage is simple: run hns, speak, press Enter, and the transcription will be shown on screen and automatically copied to the clipboard, ready to be pasted into any application.

Key Features:

• Complete Privacy: Audio is processed entirely on-device using faster-whisper. No cloud servers, no telemetry, and no data leaves your device.
• Works Offline: After the initial model download, it runs completely offline - perfect for sensitive work or areas with poor connectivity.
• Automatic Clipboard: Transcribed text is instantly copied to the clipboard for immediate pasting into any application.
• Multi-Language Support: Transcribe in any language supported by Whisper models.
• Configurable Models: Choose from multiple Whisper models (base to large-v3) to balance speed vs. accuracy.
• Developer-Friendly: CLI tool that works with terminals, IDEs, AI assistants like Claude, and any text input field.
• MIT-licensed: Fully transparent, auditable source code with no hidden functionality.

hns is perfect for:

• Developers who want voice input in terminals and code editors
• Privacy-conscious users avoiding cloud transcription services
• Quick dictation into AI assistants, Slack, emails without context switching
• Anyone who needs reliable speech-to-text without monthly fees or internet dependency