Free Malay transcription deserves a trial clip before real work. For Malay, published tests of others on the earlier large-v2 model give a word error rate of 8.7; no large-v3 figure was found for Potok. Send thirty seconds and read the result against the audio.
Transcribe a recording
Send your voice message, audio or video to @potok_now_bot on Telegram and the text comes back in the chat.
Open @potok_now_bot in Telegram
Files up to 2 GB · export to TXT, MD, DOCX, PDF, SRT and VTT · speaker labels and timestamps · about 30 languages
Key facts
- The 8.7 comes from published tests of others on large-v2, because no large-v3 figure was found; it is not our measurement.
- Malay is closely related to Indonesian, so check that spelling follows the variety you speak.
- Everyday speech often mixes in English words, which makes a good test case.
- Speaker labels, when switched on, show up as numbers.
What a transcript looks like
[00:00] Speaker 1: Mari kita mulakan mesyuarat. Pertama, kita semak keputusan minggu lalu, kemudian kita rancang langkah seterusnya.Meaning in English: Let's start the meeting. First, we'll review last week's results, then plan the next steps.
Written by us to show the script and the layout of a transcript. It is not the output of the recognition engine.
Published quality
| Language | Verdict | Published result | Notes |
|---|---|---|---|
| Malay (Bahasa Melayu) | Recommended | word error rate 8.7% (large-v2, FLEURS) Whisper paper | Published tests show a low error rate for this language. No large-v3 figure was found; the figure is from the earlier large-v2 model. |
These numbers come from published tests by other people, not from tests by Potok. Your own audio, with its noise, accents and microphone, will differ.
How to do it
- Choose a mixed clip. Use speech that has everyday vocabulary plus a few English terms, such as product names.
- Open the bot. Open @potok_now_bot in Telegram and send the clip as a voice message, an audio file or a video.
- Check borrowed words. See whether English terms were written as you would write them, in Latin letters, or turned into odd spellings.
- Check the variety. Look for Indonesian spellings in your Malay text, and note each one.
- Export and keep the best. Save the transcript as TXT, DOCX, PDF, SRT or VTT. Files up to 2 GB are accepted.
| Feature | Details |
|---|---|
| Where it works | Telegram bot @potok_now_bot; large files go through the bot's mini app |
| File size | Up to 2 GB |
| Export | TXT, MD, DOCX, PDF, SRT, VTT |
| Speakers and time | Speaker labels and timestamps can be switched on |
| Languages | About 30 languages with good published results; a few more with warnings |
| Translation | The transcript is not translated into another language |
| Dictation | A separate Potok desktop app lets you dictate text into any program; it recognises Russian speech |
Common mistakes
| Mistake | What happens | Fix |
|---|---|---|
| Sending a clip that is half another language | The tally mixes two languages and tells you little. | Keep the test mostly in Malay. |
| Assuming Malay and Indonesian always match | A few spellings differ and look wrong to a reader. | Mark them and decide if they matter. |
| Using music-covered audio | Background songs hide quiet words. | Cut a stretch of clear speech. |
| Reading only the first lines | Errors tend to pile up later, when the speaker gets tired or the room gets loud. | Read the whole sample, then compare the start with the end. |
More
See the full list of languages, the tools and the glossary.
Frequently asked questions
Is the Malay figure recent?
It is from the earlier large-v2 model, since no large-v3 figure was found.
Is 8.7 your own result?
No. It comes from published tests of others.
Will English words in a sentence be understood?
Short borrowed words often come through, but test with your own mix.
Can I turn the text into English?
No. The transcript stays in the language of the recording.
How long can a recording be?
Files up to 2 GB are accepted, and a long result arrives as a file.
Should I test more than one clip?
Yes. Two short clips in different conditions tell you more than one long clip in the same room.
Sources
Each source was opened by hand on the date shown. Pages change, so open the source if a detail matters to you.
- Word error rate (Wikipedia), Wikipedia · checked
- Whisper speech recognition system (Wikipedia), Wikipedia · checked
- Whisper paper, Table 13 (large-v2 on FLEURS) · checked
Try it on your own recording
Send your voice message, audio or video to @potok_now_bot on Telegram and the text comes back in the chat.
Open @potok_now_bot in Telegram