English is where Potok is strongest in published tests: the large-v3 model reaches a word error rate of 4 in one and 4.25 in another, and a third-party test of the turbo variant averages 4.80. Send voice notes, calls, video or files up to 2 GB to the Telegram bot.
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
- Published large-v3 word error rates for English run from 4 to 4.25, measured by others and not by us.
- A third-party test of the turbo variant averages 4.80, so even the smaller model stays close.
- Accents, jargon and cross-talk move real results more than the average does.
- Homophones such as there and their are the typical leftover mistakes.
What a transcript looks like
[00:00] Speaker 1: Let's start the meeting. First, we'll review last week's results, then plan the next steps.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 |
|---|---|---|---|
| English | Recommended | word error rate 4% (large-v3, FLEURS) Canary-1B-v2 paper, Whale paper | Published tests show a low error rate for this language. Published large-v3 results range from 4.0 to 4.25; the turbo variant averages 4.80 in a third-party test. |
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
- Pick the best source. Use the original file rather than a screen recording of a call, which adds noise and compression.
- Message the bot. Send audio, video, a forwarded voice message or a YouTube link to @potok_now_bot.
- For a feature-length file. The cap is 2 GB. The Upload button in the mini app is the way to send a very long recording.
- Turn on diarisation-style labels. Speaker labels appear as Speaker 1, Speaker 2. Add timestamps if you will quote the audio.
- Proof the specifics. Check product names, acronyms, numbers and homophones.
- Export. TXT, MD, DOCX, PDF for text; SRT or VTT for captions.
| 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 4 means four wrong words in a hundred on your file | Heavy accents and noise can push the real rate far higher. | Treat the figure as a benchmark, not a promise. |
| Missing homophones | Words that sound alike are swapped and the text still looks plausible. | Read the transcript with the sentence meaning in mind. |
| Not spelling out acronyms before the call | Unfamiliar jargon becomes ordinary words. | Keep a short term list and correct afterwards. |
| Using a call recorded through a speakerphone | Echo drops words from both sides. | Record each speaker close to the microphone. |
More
See the full list of languages, the tools and the glossary.
Frequently asked questions
How good is English transcription?
Published large-v3 tests report word error rates of 4 to 4.25 for English, among the lowest on our list. These are other groups' numbers.
Does it handle British, American and Indian accents?
The published figures do not break results down by accent. Test a short clip from your speaker if the accent is strong.
Can I transcribe a YouTube video?
Yes. Paste the link into the chat and the bot transcribes the speech.
Are speaker names filled in?
No. Labels are numbers, and you rename them yourself in the exported file.
Is the English text translated?
No. The result stays in English, and the bot does not translate it into another language.
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
- Canary-1B-v2 paper, Table 11 (Whisper large-v3 on FLEURS) · 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