Next-Generation Speech Recognition & AI Workflow for Mac
Real-time internal system audio subtitles, high-accuracy file transcription, word-level timestamps, speaker diarization, and automated AI polishing workflows.
Core Capabilities
Built for Power Users, Language Learners & Creators
Everything you need from live low-latency meeting captions to batch media subtitle generation.
Realtime Live Subtitles
Record system audio via ScreenCaptureKit or microphone in real-time. Floating translucent window with customizable typography and echo cancellation.
Media File & URL Transcription
Drag-and-drop local audio/video files or paste online video and audio streams. Batch processing with high-speed GPU acceleration.
Word Timestamps & Diarization
Precision alignment down to millisecond word-level timestamps. Multi-speaker clustering without requiring voice enrollment.
Automated AI Pipeline
Chain ASR transcription, LLM translation/proofreading/summarization, and TTS voiceover into a reusable one-click workflow template.
Privacy-First & Offline Speech
Native on-device Apple Speech recognition for offline transcription without internet access. Pair with local Ollama or OneASR for 100% data privacy.
BYOK & Service Aggregation
Seamlessly connect OpenAI, Claude, DeepSeek, OneASR, Alibaba Cloud, Tencent, and custom self-hosted endpoints with Bring-Your-Own-Key.
Modular Architecture
Visual Pipeline, Endless Possibilities
DuRT abstracts complex speech and LLM processing into 6 standardized, chainable step modules.
Audio Source
ScreenCaptureKit, Mic, File, or Web URL
ASR Recognition
Apple Native Speech, OneASR Gateway, OpenAI Whisper
LLM Polish & Translation
Proofread, Translate, Segment, Summarize
TTS Voiceover
Multi-role Voice Synthesis & Speed Tuning
Export Artifacts
SRT, VTT, JSON, TXT, Audio MP3/AAC
Elevate Your macOS Audio & Subtitle Workflow
Download DuRT today and experience the speed, accuracy, and power of native macOS speech intelligence.