Turkish speech to text: score and spelling checks

By the Potok teamUpdated ~2 min read

Turkish speech to text in Potok is a recommended language, with a published word error rate of 8.4 from the earlier large-v2 model. No large-v3 figure was found. The number is from other people's tests, not our measurement. Send voice notes, audio or video 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

  • Turkish stacks endings onto a root, so one long word can hold a whole phrase, and one wrong ending is a full word error.
  • The dotted and dotless forms of the letter i are different letters, so a swap can produce another word.
  • The 8.4 score is for large-v2, so a newer model may differ on the same audio.
  • Vowel harmony makes endings predictable, which helps you spot a wrong one by sight.

What a transcript looks like

Sample sentence in Turkish
[00:00] Speaker 1: Toplantıya başlayalım. Önce geçen haftanın sonuçlarını gözden geçireceğiz, sonra sonraki adımları planlayacağız.

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

Quality for this language
LanguageVerdictPublished resultNotes
Turkish (Türkçe)Recommendedword error rate 8.4% (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

  1. Pick the clip. A lecture, a phone note or a meeting works. Choose one where the speaker is not rushing.
  2. Share with the bot. Open @potok_now_bot and send the file, a forwarded voice message or a YouTube link.
  3. Check size. Up to 2 GB is accepted, with the Upload button for large ones.
  4. Search for the letter i. Look at words with dotted and dotless forms and compare with the sound.
  5. Read long words. Check the end of each long word, where suffixes pile up.
  6. Export. Choose TXT, MD, DOCX, PDF, SRT or VTT.
What Potok can do
FeatureDetails
Where it worksTelegram bot @potok_now_bot; large files go through the bot's mini app
File sizeUp to 2 GB
ExportTXT, MD, DOCX, PDF, SRT, VTT
Speakers and timeSpeaker labels and timestamps can be switched on
LanguagesAbout 30 languages with good published results; a few more with warnings
TranslationThe transcript is not translated into another language
DictationA separate Potok desktop app lets you dictate text into any program; it recognises Russian speech

Common mistakes

Common mistakes and how to fix them
MistakeWhat happensFix
Treating one wrong suffix as minorIt changes tense or person.Read endings with care.
Using a tool that changes letter case blindlyUppercase and lowercase of the letter i do not map as in English.Edit with a Turkish-aware editor.
Quoting 8.4 as a latest-model resultIt is a large-v2 figure.Name the model.
Skipping speaker labels in a debateTurns merge.Turn labels on.

More

See the full list of languages, the tools and the glossary.

Frequently asked questions

What is the published score for Turkish?

A word error rate of 8.4 for the large-v2 model.

Why is there no large-v3 figure?

None was found for Turkish, so we show the earlier model's number.

Why are long words a risk?

Turkish adds many endings to one word, so a single wrong ending can make the whole word count as an error.

Can I get SRT subtitles?

Yes. Switch timestamps on and export SRT or VTT.

Does the bot convert Turkish into another language?

No. The transcript stays in Turkish, in the same language as the audio you sent to the bot.

Sources

Each source was opened by hand on the date shown. Pages change, so open the source if a detail matters to you.

  1. Word error rate (Wikipedia), Wikipedia · checked
  2. FLEURS speech dataset (Google, Hugging Face), Hugging Face · checked
  3. 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