结论先行的复盘
把你的回答按「先说结论，再说理由」的结构重新排一遍，再对照原句练习。
Conclusion-first rehearsal
Your answer gets restructured "conclusion first, reasons after," so you can rehearse against a version you can actually use.
Songlet · 口语表达训练
练过的，直接用得上。
录一段回答，Songlet 帮你看清哪里在说「呃」「就是」，哪里在打太多折扣，哪里说得太空泛——然后把它练成一句你敢直接说出口的话。
Songlet · Spoken expression training
Practise here, say it for real.
Record an answer, and Songlet shows you exactly where it turns into filler, hedging or vague language — then helps you rehearse a version you can actually say out loud.
「呃，就是……我觉得这个项目大概可能还不错吧。」
"Um, so... I think this project is, like, pretty good, I guess."
填充词Filler
模糊限制Hedge
空泛表达Vague
目前还未在 App Store 上架，正在准备中。 Not yet available on the App Store — it's on the way.
它做什么
从一段录音,到一句更清楚的表达
Songlet 围绕一个简单的循环设计：说一遍，看清问题，练一遍更好的版本。
What it does
From a recording to a clearer sentence
Songlet is built around one simple loop: say it, see what's getting in the way, then rehearse a better version.
本机流式双语识别
录音与语音识别全程在你的设备上完成，支持中英文流式转写，不需要联网。
On-device streaming, bilingual
Recording and speech recognition run entirely on your device, streaming in both Chinese and English — no network needed.
可选的 AI 教练
在你明确同意之后，Songlet 才会用 Apple 智能（设备端或私密云计算）帮你生成表达报告；你可以随时在设置里关闭。
An optional AI coach
Only after you explicitly agree does Songlet use Apple Intelligence (on-device or Private Cloud Compute) to help write your report — and you can turn it off in Settings at any time.
Privacy
Your recordings stay on your phone
Speech recognition runs offline by default, and recordings and transcripts are stored on the device. No ads, no third-party tracking. Read the full privacy policy →