Overview & Installation
Welcome to DuRT (Du Speech Recognition & Transcription), a native macOS productivity application designed for high-performance speech-to-text conversion, real-time live captioning, and AI-powered audio workflow automation.
Key Highlights
- ⚡️ Real-time Live Subtitles: Capture internal macOS audio or microphone input with ultra-low latency and customizable floating captions.
- 📁 Universal Media Transcription: Transcribe local audio/video files or streaming media links with millisecond word-level timestamps and speaker diarization.
- ⛓️ Visual AI Workflows: Chain together speech recognition, large language model proofreading/translation, and text-to-speech voiceover in one click.
- 🔒 Local-First & Privacy: Native Apple on-device Speech Recognition—process audio locally without sending data to the cloud.
- 🌐 Multi-Engine Aggregation: Bring your own API keys for OpenAI, Anthropic Claude, DeepSeek, OneASR, Alibaba Cloud, Tencent, Ollama, and more.
System Requirements
- Operating System: macOS 13.0 (Ventura) or later (macOS 14 Sonoma & macOS 15 Sequoia fully supported).
- Architecture: Universal binary optimized for Apple Silicon (M1/M2/M3/M4) and Intel Macs.
- Memory: 8 GB RAM minimum.
Installation
Install directly through the official Apple Mac App Store for automatic updates and sandbox security:
Next Steps
- Check the macOS Permissions Guide to enable microphone and screen/system audio recording.
- Explore Live Caption & Subtitles for meetings and video learning.
- Learn about AI Workflow Pipelines.
