Real-time speech-to-text captioning with on-the-fly AI translation, available as native apps for Windows and macOS.
LCT listens to live audio (system playback or microphone), transcribes it, and streams a translation alongside the original — useful for following talks, meetings, videos, or calls across a language barrier.
| Platform | Stack | Status | Docs |
|---|---|---|---|
| Windows | C# / .NET 8 (WPF) + LiveCaptions | Stable | windows/README.md |
| macOS | SwiftUI + SFSpeechRecognizer + local LLM (Ollama) |
Active development | macos/README.md |
The two apps are independent codebases tuned to each platform's native speech and UI frameworks. They share this repository for common documentation, translation benchmarks, and a single issue/release home.
.
├── windows/ # Windows app (.NET solution, source, installer scripts)
├── macos/ # macOS app (Swift Package, SwiftUI)
├── benchmark/ # Cross-platform translation-quality benchmarks
├── docs/ # Shared / macOS user documentation
└── .github/ # CI workflows (per-platform, path-filtered)
- Windows — open
windows/LiveCaptionsTranslator.slnin Visual Studio, ordotnet buildinsidewindows/. See windows/README.md. - macOS —
cd macos && ./package-app.sh(orswift build). Requires macOS 15+ and a local Ollama install. See macos/README.md.
See LICENSE. Applies to both platform apps.