Transcribe Chinese oral exam practice recordings for PSLE prep

Why a transcript helps oral exam prep

A transcript makes a practice session reviewable. Instead of asking a student to remember what they said three minutes into a passage reading or a stimulus-based conversation, you both look at the same text. Vocabulary choices, sentence structure, filler words, and abandoned sentences are all visible on the page. Where the transcript comes out garbled or different from what the student meant to say, that is often a useful flag that the pronunciation was unclear and worth drilling again.

To be clear about what NexScribe does and does not do: it transcribes speech to text. It does not grade, score, or assess the student against any exam rubric. The judgement stays with the tutor or parent. The transcript just makes that judgement faster and more concrete.

How it works

1. Record the practice session. Use any phone voice recorder during the reading or conversation practice, or record directly in the browser on our homepage. MP3, WAV, M4A, and MP4 files all work, and Chinese file names are fully supported.
2. Upload it at scribe.nexsprint.ai. Choose the file, enter the email address the transcript should go to, and submit. No account, no app to install.
3. Get the transcript by email. Most recordings are processed within 5 to 15 minutes. The full text arrives in your inbox, ready to read or print.
4. Review it with the student. Go through the transcript together: highlight strong vocabulary, mark sentences that trailed off, and note the spots where the text does not match what was intended, then practise those again.

An honest note on Mandarin accuracy

English is NexScribe's most accurate language. Mandarin is supported by the underlying model along with 90+ other languages, and real tutors use NexScribe for exactly this workflow, but accuracy varies with recording quality, background noise, and accent. Since the free tier costs nothing and needs no sign-up, the sensible first step is to run one of your own practice recordings through it and decide for yourself whether the output is good enough for your review sessions.

Your recording. Your privacy.

• No third-party APIs.Audio is processed with WhisperX on our own servers, never sent to an external transcription API.
• No permanent storage.The recording is permanently deleted from all systems once your transcript is produced.
• Singapore-hosted.Processing happens on dedicated hardware in Singapore, under PDPA. See our privacy policy for the full detail.

Common questions

Is it really free?

Yes. Basic transcription is free with no account required. You upload the recording, enter an email address, and the transcript arrives in your inbox. Paid Pro plans add features like speaker labels and a transcript dashboard, but a tutor reviewing one practice recording at a time does not need them.

Does it handle Mandarin and mixed English-Chinese recordings?

The underlying speech model (WhisperX) supports over 90 languages including Mandarin, and it copes with the English-Chinese switching that is common in Singapore practice sessions. English is our most accurate language and accuracy on Mandarin varies with recording quality and accent, so run one free practice recording through it first and judge the output against your own material before relying on it.

What file formats and sizes are supported?

MP3, WAV, M4A, and MP4 files up to 50MB on the free tier (200MB on Pro). That covers recordings from any phone voice recorder app. File names in Chinese characters are fully supported, so a file named after the passage or the student uploads without any renaming.

Is my child's recording private?

The recording is processed on our own servers in Singapore using WhisperX, an open-source speech model. It is never sent to any external API or third-party cloud provider, and once the transcript is produced the audio file is permanently deleted from all systems. We operate under Singapore's PDPA.