Add audio or video
Drop one or more audio or video files, or record from your microphone. Pick the language, or let it be detected.
WAVESPEED AI SPEECH TO TEXT Your audio stays on your device
100% FREE
Free Speech to TextTranscribe audio and video, right in your browser.
Turn interviews, lectures, meetings and voice memos into text, in about 100 languages.
Play along, fix any word, then download a Word file, plain text or subtitles.
No minute limitTXT, Word, SRT & JSONStays on your device
No sign-up No upload No watermark
Drop audio or video files
Or record from your microphone. Several files are fine.
MP3, WAV, M4A, OGG, FLAC, MP4, MOV, WebM and more Recognition runs in your browser. Your recordings are never uploaded.
THREE SIMPLE STEPS
How to convert speech to text
Transcribe and check
Press Transcribe and watch the text appear. Click a paragraph to hear it, and double-click to correct a word.
Download
Copy the text, or download it as plain text, a Word document, SRT or VTT subtitles, or JSON with word times.
A transcript you can work with
Paragraphs with times
The text is split into readable paragraphs at pauses and sentence ends, each with its start time. Click one to play the recording from there.
Easy to correct
Edit a paragraph in place, split it with Enter, join it with the one before, or replace a misheard name everywhere at once. Undo works too.
Many files at once
Add a whole folder of recordings: they are transcribed one after another, and you can download every transcript in a single ZIP.
Fast or Accurate
Fast is a small download that handles clear speech well. Accurate is larger and slower and makes fewer mistakes with accents, noise and other languages.
Record and transcribe
No file? Record from your microphone and the recording is transcribed as soon as you stop.
Private by design
Recognition runs on your own device. Your recordings and transcripts are never uploaded; the transcripts are kept in this browser until you remove them.
WaveSpeed AI
Got the words? Now make the visuals.
Turn your transcript into images, clips and 3D models with WaveSpeed AI.
Image Generator
Create a cover, slide or thumbnail for your talk or podcast.
Video Generator
Turn a script or a quote from your recording into a short video.
3D Generator
Make a 3D model of a product or character you talked about.
- 1,000+ AI models
- Image, video and 3D
- Pay as you go
$1 in free credits for eligible new accounts. No card needed. Generators have separate account and credit requirements. The speech to text converter is free.
/filters:quality(82)/media/images/1791123751656780930_IfFPY7R0.webp){kind=small}
/filters:quality(82)/media/images/1791123737177812425_RjKT3cku.webp){kind=small}
/filters:quality(82)/media/images/1791123748309870128_xlkNgK8N.webp){kind=small}
/filters:quality(82)/media/images/1791123738630744263_oR09R0aj.webp){kind=small}
/filters:quality(82)/media/images/1791177033162457776_uQ09OY7g.webp){kind=small}
/filters:quality(82)/media/images/1791123734319125607_JdnwGPY7.webp){kind=small}
/filters:quality(82)/media/images/1791123746355630206_udnxGPY7.webp){kind=small}
/filters:quality(82)/media/images/1791123744771362580_yC0sPezZ.webp){kind=small}
/filters:quality(82)/media/images/1791123738606970169_5KIR1ais.webp){kind=small}
/filters:quality(82)/media/images/1791123748197655725_EUmuFNW6.webp){kind=small}
Speech to text FAQ
Is this speech to text converter free?
Yes. There is no account, no watermark and no limit on files or minutes. Recognition runs on your device, so you can use it as much as you like.
Are my recordings uploaded?
No. Your audio and video stay on your device and are transcribed in your browser. The only download is the speech model, the first time you use it; your browser keeps it for next time.
Which languages does it understand?
About 100, including English, Spanish, Chinese, Hindi, Arabic, French, German, Japanese, Portuguese and Russian. It detects the language from the first words, or you can choose it.
How accurate is it?
On clear English speech both sizes get about 95 words in 100 right. For other languages choose Accurate: in our Chinese test it got about 9 characters in 10 right, Fast only 6. Accents, noise, people talking over each other and rare names lower accuracy, so check names and numbers.
How long does it take?
It depends on your device. On a fast desktop processor, Fast needed a fifth to a third of the recording's length and Accurate two fifths to a half; laptops and phones take longer, and browsers with WebGPU can use the graphics card. You see an estimate before you start and the time left while it works.
Which files can I transcribe?
Audio such as MP3, WAV, M4A, AAC, OGG, Opus and FLAC, and videos such as MP4, MOV, WebM and MKV. Only the sound is used.
Does it add punctuation and paragraphs?
Yes. The text comes with the punctuation the speech model writes, and it is split into paragraphs at pauses and sentence ends. You can split and join paragraphs yourself.
Can it tell speakers apart?
Not yet. Long pauses usually start a new paragraph, which often follows a change of speaker; you can add names while you edit.
More free tools like this
All Free Tools
Every word, on your own device.
Free, private and unlimited.