Transcribe Audio to Text – Free Online Transcriber

A free transcriber that turns speech into text on your device

When you need to transcribe audio to text, typing it out by hand takes around four times as long as the recording itself. An automatic transcriber does the listening for you. Drop in a file, press Transcribe to Text, and the words appear as a transcript you can edit, copy into a document or save as subtitles.

Most transcription services upload your audio to their servers and charge by the minute. This one works differently. It uses Whisper, an open-source speech recognition model created by OpenAI, and runs it right in your browser. As a result, the transcription is free, there is no account, and a confidential interview or a personal voice memo stays on your computer or phone.

How to use the Audio to Text Transcriber

1. Choose your audio or video fileTap Choose File or drag a file onto the box. MP3, WAV, M4A, OGG, FLAC, AAC and the sound of MP4, MOV or WEBM videos all work, up to 500 MB.
2. Pick the spoken languageLeave Spoken language on Detect automatically, or choose it yourself, for example Spanish or Portuguese. Choosing it helps with short clips and strong accents.
3. Choose the accuracyBalanced suits clear English recordings. Best gives the most accurate text for other languages and harder audio, while Fast is the quickest on a phone.
4. Transcribe, then copy or downloadPress Transcribe to Text and watch the live preview. When it is done, edit the text if needed, then copy it or download a TXT, SRT or VTT file.

Transcribe your audio now

How the audio transcriber works

The first time you transcribe, your browser downloads the speech model you picked, between 41 MB and 249 MB, and keeps it in its cache. After that, the model loads from the cache in seconds. Your file is decoded on the page, converted to the 16 kHz mono sound the model expects, and then processed in 30-second windows that overlap slightly, so no words are lost between them.

Each window becomes a few lines of text with start and end times. The tool then joins them into paragraphs, starting a new one when the speaker pauses. Because everything happens on your device, the speed depends on your computer or phone. A modern laptop is the fastest option, and long recordings simply take a little longer.

• Fast (41 MB). The smallest model. Good for clear English speech and for phones with little memory.
• Balanced (77 MB). The default. Accurate on clean recordings in English and other widely spoken languages.
• Best (249 MB). The most accurate choice for Spanish, Portuguese, French, German and other languages, and for noisy or fast speech.

Transcribe Spanish audio to text, and 25+ other languages

You can transcribe Spanish audio to text just as easily as English. The transcriber recognises English, Spanish, Portuguese, French, German, Italian, Russian, Turkish, Arabic, Hindi, Indonesian, Japanese, Chinese, Korean and many more. By default, it listens to the first 30 seconds and detects the language for you, then shows which one it found next to the transcript.

For the best results in Spanish or Portuguese, choose the language yourself and set Accuracy to Best. This matters most for short clips, mixed languages and regional accents, such as Latin American and European Spanish, or Brazilian and European Portuguese. The text is written in the language that is spoken, so a Spanish interview gives you a Spanish transcript.

Transcribe Spanish audio now

Audio file transcription for every kind of recording

Audio file transcription is useful far beyond the office. Here are the recordings people turn into text most often.

• Interviews and research. Get a searchable transcript of a journalist interview, a user study or a podcast guest, ready to quote.
• Lectures and classes. Transcribe a recorded lecture into study notes you can search, highlight and summarise.
• Meetings and calls. Turn a Zoom, Teams or phone call recording into minutes, and keep a written record of decisions.
• Voice memos and ideas. Speak your notes on a walk and turn them into text later. You can record straight from your browser with our Voice Recorder.
• Videos and subtitles. Load an MP4, MOV or WEBM video and save SRT or VTT subtitles to add captions on YouTube or in your video editor.
• Podcasts. Publish show notes and a full transcript, which makes each episode easier to find in search engines.

Save your transcript as TXT, SRT or VTT

When the transcription is finished, the text appears in an editor. Fix a name or a technical term, then press Copy Text to paste it into Google Docs, Word or an email. You can also switch to the Timestamps view to see when each sentence was spoken.

For files, choose TXT for plain text, SRT for subtitles in YouTube, Premiere Pro, DaVinci Resolve or VLC, and VTT for web video players. The subtitle files use the timestamps from the model, so each line appears on screen while it is being said.

Tips for a more accurate transcription

Speech recognition is good, but it is not perfect. It works best on clear speech, so a few small steps make a real difference.

• Record close to the microphone. A phone held near the speaker beats a laptop across the room. Less echo and background noise means fewer mistakes.
• Choose the language and Best accuracy. For anything other than clear English, these two settings give the biggest improvement.
• Cut out music and silence. Long intros, music or silence can confuse the model. Trim them first with the MP3 Cutter.
• Check names and numbers. Proper names, brands and figures are the most common errors, so read them over before you share the transcript.
• Split very long recordings. For recordings of several hours, cut them into parts. It keeps memory use low, especially on phones.

Private, free and without limits

There is no sign-up, no watermark and no monthly quota, so you can transcribe audio to text free as often as you like. Since the speech model runs locally, your audio is never sent to a server, and nobody else can listen to it. The only download is the model itself, which comes from Hugging Face, a public library of AI models.

This transcriber is part of MP3 Converter, a set of free audio tools that all work in your browser. If your recording is a video, you can also keep just the sound with our Video to MP3 converter.

Start transcribing for free

Frequently asked questions