Convert Turkish audio to text
Turkish speech transcribed to properly written Turkish text: interviews, lectures, voice notes and video.
Free: 60 min per 30 days, files up to 30 min / 50 MB · Paid plans: 10 to 100 hrs, files up to 2 GB. Big video? Upload just its audio track: same transcript, much faster upload.
Encrypted in transit · audio & video auto-delete in 30 days · never used to train AI
Optional settings
Set this for the most accurate result on quiet or mixed audio.
Get a second copy translated into another language (Pro).
Uncommon words the AI might misspell. List them so they come out right.
Last updated 21 July 2026
Turkish has letters that careless tools mangle: the dotted and dotless i (i/ı) are different letters that change word meaning, and ğ, ş, ç, ö, ü carry real weight. Transcription here returns properly written Turkish (İstanbul, not Istanbul mangled), because the model transcribes Turkish as Turkish rather than forcing it through an English alphabet.
Turkish is agglutinative: single words carry what English says in a phrase (evlerinizden: "from your houses"), and the transcript preserves those long forms as the single words they are. Set the spoken language to Turkish before uploading; it matters more for Turkish than for English or Spanish.
Turkish audio often crosses borders before it reaches a transcript: voice notes between İstanbul and Berlin, diaspora family interviews, Turkish-language research conducted from the US. Many of those recordings mix Turkish and German; when a conversation is mostly German with Turkish phrases, transcribe it under German instead, and keep Turkish-dominant audio here. For material that needs an English version, Pro's stored translation returns the Turkish transcript and English text together, the workflow described in how to translate audio to text.
The free tier is enough to test your own audio: the free hour each month handles a week of typical WhatsApp voice notes or one full lecture, in files to half an hour, with TXT export and no card. Paid plans add DOCX/PDF/SRT/VTT exports, and Pro adds stored English translation, which is useful for diaspora family recordings and research interviews where the audience reads English.
How it works
- 1Set 'Spoken language' to Turkish under the upload box.
- 2Upload the recording: MP3, WhatsApp voice notes, calls, video.
- 3Edit inline, export, or read the English translation (Pro).
Frequently Asked Questions (FAQs)
Are ı, ğ, ş and the other Turkish letters preserved?
Yes. The transcript is written Turkish with its full alphabet, in every export format. The dotted/dotless i distinction is handled by the model itself, not patched afterwards.
Can I translate a Turkish recording to English?
On Pro, yes: Turkish transcript first, English translation second, both stored together. The two-step design means a mishear shows up as a checkable Turkish error instead of disappearing into the English.
Does it handle Turkish mixed with English or German?
Loanwords and short phrases usually survive; full sentence-switching, the Turkish-German weave common in diaspora speech, is harder for any model. Transcribe under whichever language carries the recording, and expect the switch points to need a manual glance.
Do long suffixed words come out correctly?
Yes, as the single words they are. Turkish stacks suffixes where English uses whole phrases, and the transcript writes gelemeyeceğim as one word rather than a broken-up phrase. Vowel harmony makes most forms predictable for the model; the rare heavily stacked form in fast speech is worth a glance in the editor, one click per fix.
Is Azerbaijani supported under the Turkish setting?
No. Azerbaijani is Turkish's closest major relative and speakers partly understand each other, but it is a separate language with its own letters (ə, q, x) and sound patterns, and the Turkish model will mangle it. Turkish from any region, Cypriot accents included, belongs here; Azerbaijani does not.
Is there a free tier for Turkish?
Yes. The free hour renews every month and covers a steady flow of voice notes or one long interview, in files to 30 minutes, with TXT export and no card. Judge it on your own audio first.
An honest note: Turkish support is solid but younger in AI transcription than English or Spanish; noisy multi-speaker recordings need a heavier editing pass. Clear, close-mic speech does well; test on the free plan before a large batch.
Related pages
Arabic transcription, honestly rated
Voice to TextVoice notes and memos to text
How to Translate Audio to TextThe two-step workflow
Explore everything in Languages or browse all transcription tools.