Potok transcribes Spanish voice messages, audio and video through a Telegram bot. Published tests of the speech model behind it report a word error rate of 3.12% for Spanish, one of the lowest on our list. Files up to 2 GB work, and you can export TXT, MD, DOCX, PDF, SRT and VTT.
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 3.12% figure comes from other people's published tests, not from our own measurements.
- Those tests give one result for Spanish in general. Accents from Spain, Mexico and Argentina are not reported separately.
- Accent marks change meaning in Spanish: él is a pronoun and el is an article, so check them before you publish.
- The text stays in Spanish and is not translated into another language.
What a Spanish transcript looks like
A transcript is plain text. With speaker labels and timestamps switched on, each passage starts with a time and a speaker number, as in the sample below.
[00:00] Speaker 1: Vamos a empezar la reunión. Primero repasaremos los resultados de la semana pasada y después planificaremos los próximos pasos.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.
Speaker names are not guessed: you get "Speaker 1" and "Speaker 2", and you rename them yourself in the exported file.
How good is Spanish transcription?
Quality is reported as a word error rate: the share of words that differ from a reference text written by a person. Lower is better. We have not run our own tests for this page, so the table uses published results.
| Language | Verdict | Published result | Notes |
|---|---|---|---|
| Spanish (Español) | Recommended | word error rate 3.12% (large-v3, FLEURS) Canary-1B-v2 paper | Published tests show a low error rate for this language. One of the best published results; regional accents (Spain, Mexico, Argentina and others) were not reported separately. |
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.
Because the tests report one figure for Spanish in general, a recording from Spain and one from Argentina may behave differently. Test a short file with your own speakers before you send a long one.
How to transcribe a Spanish recording
- Open the bot. Open @potok_now_bot in Telegram.
- Send the recording. Send the voice message, audio file, video or YouTube link. For a large file, up to 2 GB, use the Upload button of the bot's mini app.
- Switch on speaker labels and timestamps. Do this for meetings, interviews and any call with more than one voice.
- Check names, numbers and accents. Read the Spanish text that comes back and correct names, figures and accent marks.
- Export. Save as TXT, MD, DOCX, PDF, SRT or VTT. Use SRT or VTT for Spanish subtitles.
What you get
Short transcripts arrive as text in the chat, long ones as a file. Export the text as TXT or MD for notes, DOCX or PDF to share a document, and SRT or VTT to add Spanish subtitles to a video.
| 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 |
|---|---|---|
| Assuming one figure covers every Spanish accent | A recording from Spain and a recording from Argentina may behave differently. | Send a short clip from your own speakers first and read the result. |
| Ignoring accent marks and the signs ¿ and ¡ | A missing accent can turn one word into another, for example él and el. | Search the text for the words that matter to you and correct them in your editor. |
| Expecting an English text back | The transcript stays in Spanish. The bot does not translate it. | Use a separate translation step if you need another language. |
| Using a noisy or echoing recording | Noise, echo and several people talking at once raise the error rate in every language. | Record close to the speaker, or send the cleanest version of the file you have. |
Check the result before you use it
Automatic transcripts of any language make mistakes with names, numbers and overlapping voices. Listen to the places that matter, such as quotes and figures. If the text will be published or used officially, have a Spanish speaker read it. Compare other languages on the full list of languages, or read about Japanese speech to text to see how a language without spaces is measured. For other file types, start with MP3 to text.
Frequently asked questions
Does Potok transcribe Spanish from Latin America as well as from Spain?
The published tests give one figure for Spanish in general and do not report accents from Spain, Mexico, Argentina or other countries separately. Send a short recording from your own speakers first and check the text.
Can I transcribe a Spanish video, not only audio?
Yes. The bot transcribes audio and video, so you can send a video file or a YouTube link. Files up to 2 GB are supported.
Will the Spanish transcript be translated into English?
No. The transcript stays in the language of the recording. Potok does not translate transcripts into other languages.
Which formats can I export a Spanish transcript to?
TXT, MD, DOCX, PDF, SRT and VTT. SRT and VTT are subtitle files with a time for every line.
How good is Spanish transcription?
In published tests of the speech model the word error rate for Spanish is 3.12%. That is a result from other people's tests, not our own measurement. Noise, echo and several people talking at once raise the error rate.
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
- FLEURS speech dataset (Google, Hugging Face), Hugging Face · checked
- Canary-1B-v2 paper, Table 11 (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