macOS 15.0 或更高版本，Apple Silicon（M 系列）芯片。推荐 16GB 内存。
模型已内置，安装即用，无需下载和配置；正常使用无需网络，推理完全在本机运行。
用于实时识别您听到/说出的内容。音频只在本机处理，不上传。
macOS 通过 ScreenCaptureKit 提供系统音频采集，系统将其归类在「屏幕录制与系统录音」权限下。TransLive 只取声音，不录制画面。
在「术语表」中添加课程术语的指定译法，同时这些词会作为识别热词，显著改善专有名词的识别与翻译。
全部在本机。历史记录可在「历史记录 → 在访达中显示」查看，随时可删。
问题反馈：bbzlkcn@gmail.com
macOS 15.0 or later on Apple Silicon (M-series). 16GB RAM recommended.
Models are bundled — the app works immediately after install with no downloads or configuration. Normal use needs no network, and inference stays on-device.
Q: Why the microphone permission?
For real-time on-device recognition. Audio is processed locally and never uploaded.
Q: Why does System Audio need Screen Recording permission?
macOS exposes system-audio capture through ScreenCaptureKit, which is grouped under the Screen & System Audio Recording permission. TransLive captures audio only — never your screen.
Q: How can I improve translation quality?
Add course terminology to the Glossary; those words also act as recognition hotwords.
Q: Where is my data?
Entirely on your Mac. See History → Show in Finder; delete anytime.
Email bbzlkcn@gmail.com