To try a free Catalan transcription tool, send a short clip with place names and read the text against the audio. With Potok, published tests of the earlier large-v2 model give a word error rate of 7.3 for Catalan; no large-v3 figure was found, so the number is dated.
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 7.3 is a word error rate from a published test of large-v2, not our measurement.
- No large-v3 figure was found for Catalan, so the evidence rests on the earlier model.
- Place names, street names and borrowed words are good probes for a short test.
- The transcript stays in Catalan and is not translated.
What a transcript looks like
[00:00] Speaker 1: Comencem la reunió. Primer repassarem els resultats de la setmana passada i després planificarem els propers passos.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 |
|---|---|---|---|
| Catalan (Català) | Recommended | word error rate 7.3% (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
- Record a local-flavoured clip. Read a paragraph that mentions a few towns, streets or surnames. These are harder than everyday vocabulary.
- Open the bot. Open @potok_now_bot in Telegram and send the clip as a voice message or an audio file.
- Check accents and apostrophes. Catalan text uses accent marks and apostrophes. See whether the tool keeps them.
- Note what the tool did with names. Write down every place name that came out spelled differently from the usual form.
- Scale up. If the sample is good, send the long file. Up to 2 GB is accepted, with Upload in the mini app for large ones.
| 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 |
|---|---|---|
| Judging only by common words | Everyday vocabulary hides trouble with names. | Put at least five proper names in the sample. |
| Overlooking missing accent marks | A word without its accent can read as another word. | Scan the text for accents and compare with the audio. |
| Reading 7.3 as current model quality | It was measured on the earlier large-v2 model. | Say so when you quote it, and verify with a clip. |
| Using audio with mixed languages | Switches between Catalan and another tongue confuse the check. | Test pure Catalan first, then the mixed audio. |
More
See the full list of languages, the tools and the glossary.
Frequently asked questions
How good is Catalan in published tests?
The published word error rate is 7.3, from large-v2, which puts Catalan among the good results.
Why no large-v3 number?
None was found when we gathered the evidence, so the page cites the earlier model and says so.
Can I send a video of a talk?
Yes. Video files, audio and YouTube links are accepted, within the 2 GB limit.
What formats can I save?
TXT, MD, DOCX, PDF, SRT and VTT are available for export.
Will the result switch to Spanish or English?
No. The transcript stays in the language of the recording.
Sources
Each source was opened by hand on the date shown. Pages change, so open the source if a detail matters to you.
- FLEURS speech dataset (Google, Hugging Face), Hugging Face · checked
- Word error rate (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