Transcribe an MP3 file
Choose an MP3 file and AI (Whisper or SenseVoice) transcribes it inside your browser. The audio never reaches a server. No account, no payment, no length limit.
You can normally leave this on Auto. Change it only if the result clearly has the wrong number of speakers
No file? Record with your microphone. Transcription starts after you stop recording (up to about 000 min on this device with the selected model). Your recording is not sent to any server either.
How to transcribe an MP3
- Choose the language. Auto-detect works, but picking the actual language (English for an English recording, for example) reduces misrecognition. If two or more people are talking, turn on speaker diarization as well.
- Drop the MP3 file here or click to choose it. Transcription starts as soon as the file is loaded, and lines appear in order as they are finalized.
- When it finishes, click a line to replay the audio from that point and correct any mistakes. Then choose TXT, Word or another format under "Download", or copy the text directly.
What matters when working with MP3
MP3 is a compressed format, so the bitrate of the original recording directly affects accuracy. At 128 kbps or above, ordinary conversation is recognised reliably. Below 64 kbps — or with phone recordings where the high frequencies have been stripped — you will see more dropped words. For languages other than Japanese, Chinese and Korean, switching to the small model or larger usually helps.
Everything runs on your own device. The file is not uploaded and the resulting text is not sent anywhere either, so recordings you cannot share externally — meetings, interviews — can be processed as they are.
Silent stretches are not processed, so the time it takes depends on how long people actually speak, not on the length of the file. Lectures with long pauses, or recordings where nobody talks for long periods, finish faster than their length suggests.
If something goes wrong
- If the same name or term is transcribed wrongly again and again, fix every occurrence at once with "Find & replace" on the results screen.
- In podcasts with music or jingles, sentences nobody said may appear during the music. Delete those lines on the results screen.
Frequently asked questions
Can I transcribe formats other than MP3?
Yes. WAV, M4A, AAC, FLAC, OGG, Opus, AIFF, WMA, AMR and WebM are all supported. If you pick a video file, only the audio track is extracted and processed.
Is there a file size limit?
No. There is no duration limit either. How long a recording can be processed at once depends on your device's memory and the selected model; if it does not fit, the page tells you before processing starts.
Is my audio uploaded anywhere?
No. Transcription runs inside your browser, and neither the audio file nor the resulting text is sent to any external server. Processing statistics used to fix problems (device specs, error type and similar) are sent to this site's own server, but they contain no audio, transcript text, file names or IP addresses. You can stop this by turning off the switch under “Help us improve the service” at the bottom of the page.
How long does transcription take?
It depends on your device and the model size. With WebGPU available it is often faster than real time; without it, expect several times the length of the audio.
Getting the audio ready, and after transcribing
Help us improve the service
Sending us the circumstances of a run — successful or failed — helps us track down problems (a failure rate needs both). Knowing which devices and settings fail most often lets us revise the recommended defaults and fix the bugs behind them. If you don't mind, leaving this on is a real help.
What is sent
Your device specs (GPU model, memory, core count), browser and OS, the model and language you chose, your speaker diarization setting (on or off, and the number of speakers you set), the page's display language, how far the run got (including when you stopped it or closed the page), the error type and details, the file's format with a rough size and length, and how long it took. A temporary number that links runs in the same tab (it is deleted when you close the tab). On the result screen: how many times you used editing (such as renaming speakers or splitting lines), translation, summaries and subtitle imports, the formats you chose under “Download”, how often you used “Copy text” and “Download all”, and the translation language pair.
What is never sent
Audio, transcripts (including speaker names and translations), file names, and IP addresses. The data goes only to this site's own server, never to a third-party service.