Turn Korean meetings, interviews and lectures into searchable Hangul text with speaker labels and word-level timestamps.
iOS and Android. 30 minutes free, no card.
Does SozAI transcribe Korean?
SozAI transcribes Korean audio and video into Hangul text with automatic speaker labels and word-level timestamps. It writes the speech level that was actually used rather than normalising every sentence to one register, and exports to TXT, DOCX, PDF, SRT and VTT. Recordings are processed in EU data centers and encrypted at rest, and every account gets 30 minutes a month without a card.
한국어, written the way it is actually written.
Specification
| Native name | 한국어 |
|---|---|
| ISO 639-1 code | ko |
| Writing system | Hangul |
| Text direction | Left to right |
| Speakers worldwide | ~82 million |
| Varieties covered | Seoul standard |
| Speaker labels | Yes — automatic diarization |
| Timestamps | Word level |
| Summaries | Yes, in the same language |
| Export formats | TXT, DOCX, PDF, SRT, VTT |
| Maximum file size | 500 MB |
| Free tier | 30 minutes per month, no card |
| Where data is stored | EU data centers, AES-256 at rest |
| Public transcripts in this language | 332 |
Workflow
Record straight into the app, upload a file up to 500 MB, or paste a YouTube link. SozAI reads m4a, mp3, wav, aac, mp4 and mov.
Choose Korean, or leave detection on. Recordings that mix Korean with English business vocabulary stay in a single transcript.
The transcript arrives with speaker labels and word timestamps. Search it, ask it questions, generate a summary in Korean, or export SRT and VTT subtitles.
Field notes
Korean encodes the relationship between speakers in the verb ending, and the same statement takes a different form depending on who is being addressed. A transcript that normalises everything to one speech level erases the hierarchy the meeting actually ran on, which is exactly the information a meeting record is supposed to preserve.
Korean 띄어쓰기 decides where a phrase is one word and where it is two, and the rules are intricate enough that fluent writers disagree. Spacing changes meaning in real cases. Machine transcripts of Korean tend to be judged on spacing before anything else, because it is the most visible thing to get wrong.
What is written and what is spoken diverge systematically: final consonants change under the influence of the next sound, so the audio does not contain the letters the orthography requires. Korean transcription therefore has to run the phonological rules backwards to recover the spelling.
A large part of the formal vocabulary comes from Chinese roots, and many of those words are homophones in modern Korean. Context is the only separator — the audio genuinely does not distinguish them. This is where residual errors concentrate in legal, medical and academic recordings.
From the archive
332 public Korean transcripts so far — every one produced by the same engine described above. Read them in full, no account needed.
![[Poe 3.29] 엘리 화염구빌드 제작가이드 / 매물이 없다면 만들어보세요 — Transcript — Korean transcript](https://i.ytimg.com/vi/l2P7-4gCCs8/maxresdefault.jpg)
[Poe 3.29] 엘리 화염구빌드 제작가이드 / 매물이 없다면 만들어보세요 — Transcript
인류 이전의 존재… 아눈나키조차 두려워했던 그들은 누구인가? — Transcript
YouTube Video — Transcript
08/13/2026 Lunch Broadcasting – Live stream — Transcript
다이어트 하는 법 그만 검색하고 이거 한번 보세요 — Transcript
발달장애 · 발달지연아동 조기개입 국가 지원 강화를 위한 국제심포지엄 — TranscriptSozAI transcribes over 100 languages. These have a page of their own.
Answers
Short answers you can quote. Longer ones live in the sections above.
Yes. The transcript records the speech level actually used rather than rewriting every sentence into one register. In a meeting record this is substantive information: it shows who was deferring to whom, and normalising it away would change what the document says.
Yes. Diarization runs on the audio, so each speaker gets a label and every turn a timestamp regardless of language. For interviews and recorded meetings this is what makes the transcript usable as a record you can quote from.
The transcript is written with standard 띄어쓰기. Spacing is the most visible marker of transcript quality in Korean and the rules are genuinely intricate, so a proofreading pass on long documents is still worth budgeting for.
SozAI exports SRT and VTT in UTF-8 with word-level timing, so Hangul loads correctly in YouTube, Premiere Pro, Final Cut Pro and DaVinci Resolve without an encoding step.
Every account gets 30 minutes per month free without a card. Beyond that SozAI is a flat monthly subscription with a shared pool of minutes, and the price is the same for every language. Current plans are listed on the pricing page.
Where to use it
Everything below works in Korean, with the same speaker labels and word timestamps.
Transcribe your first Korean recording
Record, upload a file or paste a link. Hangul with speakers and timestamps comes back.
Free on iOS and Android. No account required for YouTube transcripts.
Point your camera
Scan the code to open the download page on iOS or Android. 30 minutes of transcription free, no card.
