Free Spanish transcription is simple to test and worth testing by region. Published tests of others report a large-v3 word error rate of 3.12 for Spanish, one of the best, and Potok uses that model. Try clips from each accent you handle, because the tests did not split them.
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 is among the best published results, and regional accents were not reported separately.
- Spanish uses opening question and exclamation marks, which are easy to forget in a transcript.
- Speech from Spain, Mexico and Argentina can differ in rhythm and in how consonants are pronounced.
- A low score on a benchmark does not exclude errors in names.
What a transcript looks like
[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.
Published quality
| 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.
How to do it
- Collect a clip per accent. If you work with several regions, take thirty seconds from each, and keep them in separate files.
- Open the bot. Open @potok_now_bot in Telegram and send them one by one, as voice messages, audio files or videos.
- Check opening marks. Look for the inverted question and exclamation marks at the start of sentences.
- Compare accents. Line the results up and see whether one region has more wrong words.
- Export the final run. Save the text as TXT, DOCX, PDF, SRT or VTT, and for large files up to 2 GB use the Upload button.
| 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 |
|---|---|---|
| Testing in a single accent | One region does not stand for the rest. | Use at least two different ones. |
| Dropping accent marks | A dropped accent mark can turn a verb into a noun. | Proofread marked vowels. |
| Trusting a very low published number blindly | It reflects test data, not your noisy call. | Always run your own sample. |
| Forgetting names from other languages | Brand and place names get mangled. | Keep a short list to compare. |
More
See the full list of languages, the tools and the glossary.
Frequently asked questions
Is Spanish among the strongest languages?
Published tests of others give it one of the best large-v3 error rates.
Does the figure cover Latin America and Spain?
It does not separate accents, so regional results can differ.
Are inverted question marks added?
Often they are, but check a sample, as they are a common small gap.
Can I get English text?
No. The transcript stays in the language of the recording.
What size can the file be?
Files up to 2 GB are accepted, and a long result arrives as a file.
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