- 1Choose
- 2Transcribe
- 3Result
Drop your audio file
or click to browse · max 25 MB · 30 min
MP3WAVM4AAACOGG
The transcription is done by a secure external AI service. Your audio is sent there for the duration of processing then deleted, and is never kept on our servers. Learn more
How to transcribe audio in 3 steps
Add your audio or video
Pick an audio or video file — a meeting, an interview, a voice note. It is prepared in your browser before the transcription runs.
Read, correct, reread only what matters
Each segment keeps its timestamp. Passages the model was unsure about are flagged in amber; fix a word in place and the flag clears. Nothing is invented — an unmarked passage can still be wrong, and the tool says so.
Export the way you need it
Download plain text, subtitles (SRT) synced to the audio, a searchable PDF, or a Word document — with or without timestamps, your choice.
Why use this transcription tool
- Every line keeps the exact moment it was said. Jump to a passage, and the subtitles you export stay in sync with the audio.
- Passages the model hesitated on are flagged so you reread only those. It never guesses who is speaking — no invented speaker names.
- One transcription, four files. Vector PDF you can search and copy, SRT subtitles for video, and a Word document when you need to keep editing.
- Record on the go, transcribe on the same device. Long recordings are handled without freezing the page.
Frequently asked questions
Is this transcription tool free?
- Yes. You can transcribe, correct and export to text, SRT, PDF or Word with no account and no watermark.
Does it add timestamps and subtitles?
- Yes. Each segment keeps its exact moment, so you can export SRT subtitles that stay in sync with your video, or a text with timestamps in the margin.
Can it tell who is speaking?
- No, and it won't pretend to. There is no honest way to guess speakers from audio alone, so the tool never labels “Speaker 1” — it only transcribes what was said.
How accurate is it?
- It is accurate on clear speech, and it flags the passages where it hesitated so you can reread them. An unmarked passage can still be wrong — the tool says this plainly rather than pretending to be perfect.
What languages and files does it handle?
- It transcribes common audio and video files. When a language can't be laid out cleanly in the PDF, the Word export takes over so nothing is lost.