Potok turns Japanese speech into text through a Telegram bot. Published tests of the speech model behind it report a character error rate of 4.5% for Japanese. Japanese has no spaces between words, so quality is counted in characters. Files up to 2 GB work, with export to TXT, DOCX, SRT and more.
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
- Japanese quality is counted in characters, not words, because the written language has no spaces between words.
- A character error rate and a word error rate use different units, so 4.5% for Japanese cannot be compared directly with a figure for Spanish.
- Many Japanese words sound alike but are written with different characters, so names need a check against the audio.
- The text stays in Japanese and is not translated.
How Japanese quality is measured
For English or Spanish, quality is a word error rate. Japanese has no spaces to split words, so published tests count characters instead. The character error rate is the share of characters that differ from a reference text written by a person. Lower is better.
| Language | Verdict | Published result | Notes |
|---|---|---|---|
| Japanese (日本語) | Recommended | character error rate 4.5% (large-v3, FLEURS) Whale paper | Published tests show a low error rate for this language. Japanese is written without spaces, so quality is reported as a character error rate, not a word error rate. |
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.
We have not run our own tests for this page. The figure comes from a published comparison, and your own recordings may do better or worse.
A sample sentence
The sentence below shows how Japanese text appears in a transcript, with a time and a speaker number in front of it.
[00:00] Speaker 1: 会議を始めましょう。まず先週の結果を確認し、そのあと次のステップを計画します。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.
Where Japanese transcripts need a second look
Spoken Japanese has many words that sound the same but are written with different characters. A transcript can contain the right sound written with the wrong characters. Names of people and companies are the most common case. Read those parts against the audio, and keep a list of the names that appear in your recordings.
Background noise and several people speaking at once make mistakes more likely in every language, including Japanese.
How to send a Japanese recording
- Open the bot. Open @potok_now_bot in Telegram.
- Send the recording. Send the voice message, audio, video or YouTube link. For a large file, up to 2 GB, use the Upload button of the bot's mini app.
- Check the names. Read the returned text and compare names of people and companies with the audio.
- Export. Save as TXT, MD, DOCX or PDF for reading, or SRT or VTT for subtitles.
What you get
Short transcripts arrive as text in the chat and long ones as a file.
| 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 |
|---|---|---|
| Comparing a character error rate with a word error rate | Japanese is counted in characters and many other languages in words, so the numbers measure different things. | Compare Japanese with Japanese figures from the same kind of test. |
| Trusting names written in kanji | The right sound can be written with the wrong characters. | Keep a list of the names in your recordings and check every one against the audio. |
| Searching the transcript by words separated by spaces | Japanese text has no spaces between words, so a space-based search finds little. | Search for the characters or the phrase you need. |
| Publishing subtitles without reading them | Long Japanese lines may not fit the screen, and a wrong character is easy to miss. | Play the video with the subtitles on and shorten lines in your video editor. |
For subtitles, read what an SRT file is, and what speaker diarization means when a transcript marks who is speaking. Not sure whether your file fits? Use the audio file size calculator. To compare other languages, see the full list of languages or the page on Spanish transcription.
Frequently asked questions
Why is Japanese quality measured in characters?
Japanese is written without spaces between words, so counting wrong words is unreliable. Published tests count the share of wrong characters instead, called the character error rate.
How good is Japanese speech to text?
In published tests of the speech model the character error rate for Japanese is 4.5%. That is a result from other people's tests, not our own measurement. Your recording may do better or worse.
Can I get Japanese subtitles as an SRT file?
Yes. Export the transcript as SRT or VTT and the text comes with a time for every line. Read the file before you publish it.
Does Potok translate Japanese speech into English?
No. The text stays in Japanese. Potok does not translate transcripts into other languages.
Sources
Each source was opened by hand on the date shown. Pages change, so open the source if a detail matters to you.
- Japanese writing system (Wikipedia), Wikipedia · checked
- Word error rate (Wikipedia), Wikipedia · checked
- Whale paper, Table 5 (Whisper large-v3 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