본문으로 건너뛰기

음성 텍스트 변환 — 무료, 가입 없이 브라우저에서

영어, 독일어, 프랑스어, 스페인어, 이탈리아어, 네덜란드어, 폴란드어, 중국어 음성을 텍스트로 변환하세요. 가입이나 오디오 업로드 없이 브라우저에서 처리합니다.

No sign-up · Runs in your browser · Audio stays on your device

Turn a recording into editable text. Select the spoken language, choose a local audio file or record your voice, then transcribe on your own device.

Whistle · About 17 MB of model files for these seven languages. Runs on your CPU.

Or drag and drop an audio file here

Up to 5 minutes and 50 MB. MP3, WAV, M4A and other formats your browser can decode.

Check names, numbers and punctuation. Recognition can make mistakes, especially with noise, accents or speech across chunk boundaries.

사용 방법

  1. Turn a recording into editable text. Select the spoken language, choose a local audio file or record your voice, then transcribe on your own device.
  2. Transcribe audio — The first run downloads model files. Keep this tab open; cached files can be reused when your browser allows it.
  3. Transcript ready. You can edit, copy or download it.

예제

Hello, this is a short recording. → Hello, this is a short recording.

세부 정보 및 제한 사항

Up to 5 minutes and 50 MB. MP3, WAV, M4A and other formats your browser can decode.

Long recordings are split locally into segments shorter than 30 seconds. Processing speed depends on your device.

Check names, numbers and punctuation. Recognition can make mistakes, especially with noise, accents or speech across chunk boundaries.

Only the selected model is downloaded from this site. Audio and transcript contents are not sent to an inference server. Site analytics and advertising still make their own requests.

FAQ

Do I need an account?

No. Select a file or use the microphone and start transcribing without registration.

Does it work offline?

Inference runs locally. Opening the site and the first model download require a connection; browser caches may be cleared. This is not a fully offline installed app.

형식 참고 자료

도구 페이지 업데이트: