Is the output in Hangul with correct spacing?

Yes. The model writes standard Hangul with spacing, not romanised text, so the transcript can be pasted straight into a document or subtitle file.

What about English words mixed into Korean speech?

They are transcribed in Hangul the way they are usually written in Korean. Long English passages inside a Korean recording may come out inconsistent, because the model follows the language you selected.

Can I make subtitles for a Korean video?

Yes. Use the Subtitle Generator page to export an .SRT or .VTT with timestamps, or translate the subtitles to English in the same run.

Does it separate the speakers?

No. The transcript is one continuous text without speaker names, so a dialogue arrives as a single block. Add speaker labels while you edit; the punctuation and sentence breaks make the turns easy to spot.

Which model is best for Korean?

Accurate is a good default for clear speech such as lectures and news. Use Best for casual conversation, slang, overlapping voices or noisy recordings, and Fast only when you need a rough draft quickly.

Transcribe Korean Audio to Text — Free, No Upload

Convert Korean audio or video into Hangul text in your browser. 100% free, no per-minute limit, no account — nothing is uploaded.

Model

Spoken

Output

Timestamps [0:05]

⬇️ One-time download: 104 MB (Accurate model 73 MB + 31 MB audio engine, shared with the video tools) — only the first time: it stays saved in your browser, so next time it starts right away. (Translating or switching model needs its own file; a browser can clear this cache if storage runs very low.) Processing takes about 4-8 min per 10 minutes of audio on a typical laptop.

What exactly is downloaded — and is it safe?

• The speech model is OpenAI's Whisper, open source, converted to run in browsers. The audio engine is FFmpeg, the open-source tool behind most video software.
• They come from Hugging Face and unpkg — public repositories used by researchers and developers worldwide, not from us.
• They are not programs. Nothing is installed, no permission is asked, and your browser keeps them in its own cache exactly like it keeps a website's images.
• They can't read your files or touch anything outside this page. Remove them any time by clearing this site's data in your browser.
• Your audio never goes the other way: this site has no server receiving files.

Drop your video or audio here or click to choose

transcribed on your device by OpenAI's Whisper — your recording is not uploaded

ℹ️ Use a recording on your device that you own or have permission to process. This tool cannot fetch YouTube or social video links. Start with a short clip. Larger models can improve accuracy but require more memory and download data. Processing time varies by device, model and recording length. Always review the result.

Korean lectures, interviews, drama scripts, meeting recordings and podcast episodes end up on this page for one reason: someone needs the words on paper, in Hangul, without paying a per-minute transcription service. The page opens with Korean already selected, which stops a recording from being mistaken for Japanese or Chinese when the first seconds are music or noise, and the output is standard Hangul with word spacing and punctuation, not a romanised approximation. Formal and casual registers both transcribe, and English loanwords are written in Hangul the way Korean speakers write them.

The work is done by OpenAI's Whisper running inside your browser in Private mode. The engine and the speech model are downloaded on first use and kept where possible; after that the audio is processed on your device, which means a confidential interview or a student's recorded lecture stays private. Fast gives a quick draft, Accurate tightens it, and Best is the choice for fast conversational speech, heavy slang or a recording with several voices. The text can be copied straight into a document or downloaded as TXT, and subtitle exports with timestamps are available for video.

Two limits to know. The transcript does not label speakers, so a two-person interview arrives as one continuous flow that you separate when editing. And numbers, proper names and company names are where errors cluster; read those lines twice.

More sizes & options

Frequently asked questions

Is the output in Hangul with correct spacing?: Yes. The model writes standard Hangul with spacing, not romanised text, so the transcript can be pasted straight into a document or subtitle file.
What about English words mixed into Korean speech?: They are transcribed in Hangul the way they are usually written in Korean. Long English passages inside a Korean recording may come out inconsistent, because the model follows the language you selected.
Can I make subtitles for a Korean video?: Yes. Use the Subtitle Generator page to export an .SRT or .VTT with timestamps, or translate the subtitles to English in the same run.
Does it separate the speakers?: No. The transcript is one continuous text without speaker names, so a dialogue arrives as a single block. Add speaker labels while you edit; the punctuation and sentence breaks make the turns easy to spot.
Which model is best for Korean?: Accurate is a good default for clear speech such as lectures and news. Use Best for casual conversation, slang, overlapping voices or noisy recordings, and Fast only when you need a rough draft quickly.

Guides

Related tools

🔒 Private mode (the default): your recording never leaves your device. Instant mode is optional, you switch it on yourself, and it sends the audio to the browser's speech service.