You have an MP3: an interview, a lecture, a voice memo, a podcast episode. You need what's in it as text, so you can quote it, search it, or turn it into notes. And you'd rather not pay for a service, upload a private recording, or sign up for a trial that asks for a card.
There are several free ways to do it, and they're not equally good. This guide walks through the options, from typing it yourself to letting a speech model do it on your own computer, and shows how to get the cleanest transcript from whichever you choose. The fastest free route is Spokly's audio and video to text tool, which runs in your browser, so the file never leaves your device.
Before you start: what affects accuracy
Whatever method you use, three things decide how good the transcript will be:
- Recording quality. A clear voice close to the microphone transcribes far better than a distant speaker in an echoing room. Background music and several people talking at once make it harder.
- The language. Tell the tool which language is spoken. "Auto-detect" works for long, clear recordings, but picking the language avoids a wrong guess, especially for short clips or ones that start with music.
- Accents and terms. Names, brands and technical words are where every method makes mistakes. Plan to check them.
None of these depend on the file format. MP3, M4A, WAV, OGG and OPUS all hold the same kind of audio; the format matters only for whether a tool can open it.
Option 1: on-device speech recognition (free and private)
Modern browsers can run a speech recognition model directly on your computer or phone. The audio is processed locally, so nothing is uploaded, and there's no limit on how many files you transcribe.
With Spokly's audio and video to text tool:
- Open the tool and drop your MP3 into the box. MP4, MOV, M4A, WAV, OGG, OPUS and WebM work too.
- Choose the spoken language.
- Choose Fast or Accurate. Fast is quicker and fine for clear English. Accurate makes fewer mistakes with accents, noise and other languages; on a computer, it's chosen for you when the language isn't English.
- Press Transcribe for free and keep the tab open while it works.
- Read the text, fix anything that's wrong, then Copy it or Download TXT.
The first time, the browser downloads the speech model once: 125 to 200 MB for Fast and 385 to 565 MB for Accurate, depending on your browser (the tool shows the exact size). After that, it starts right away and even works offline.
Limits: up to 60 minutes per file on a computer and 15 minutes on a phone. For a longer recording, split it into parts and transcribe each one.
Good for: interviews, lectures, meetings, voice memos and podcasts, especially private ones.
Option 2: type it yourself, with playback tricks
For a short recording, or one where accuracy has to be perfect, typing it yourself is still a real option, and a few tricks make it much faster:
- Slow it down. Most media players can play at 0.75× speed, which is often enough to type along without stopping.
- Use keyboard shortcuts to pause and rewind a few seconds without leaving your text editor. Many players support a global pause key.
- Type the gist first, then fill in. One fast pass for the structure, a second pass for exact words.
Typing is slow: plan on several minutes of work for every minute of audio. A good compromise is to let a speech model do the first draft and then correct it by ear.
Option 3: play the file into voice typing
Voice typing tools listen to the microphone, not to files. You can still use one by playing the recording out loud near the microphone while voice typing is on. It works, roughly, but the result is usually worse than transcribing the file directly: the speaker adds noise, the room adds echo, and any sound in the room ends up in the text.
Use this only as a last resort. If you can open the file in a transcription tool, do that instead. Voice typing is great for what it's made for: speaking new text, live.
Option 4: paid services, when they're worth it
Paid transcription services, human or automatic, make sense in a few cases: you need speaker labels for a long panel discussion, the recording is very poor, or the transcript is for legal or medical use where every word must be verified. For everyday notes, quotes and drafts, a free on-device transcript plus a careful read-through is usually enough.
How to get a cleaner transcript
- Choose the language instead of Auto-detect when you know it.
- Use Accurate for accents, noisy recordings and languages other than English.
- Cut long silences and music from the start of the file if you can. Some models guess the language from the first seconds.
- Split very long files into parts at natural breaks. Shorter files are also easier to check.
- Read it through once while listening at normal speed. Mark names and numbers as you go.
For more on the settings and what each one changes, see our guide to transcribing audio for free without uploading.
What to do with the transcript
A raw transcript is a starting point. Some common next steps:
- Notes and summaries: keep the key points and quotes, drop the filler.
- Articles and posts: a podcast episode or a talk can become an article. Our guide on turning a podcast episode into a blog post shows how.
- Subtitles: if you need each line with its start and end time, make a subtitle file from the recording instead of plain text.
- Search: a transcript makes a long recording searchable. Find the minute where something was said, then listen to just that part.
Common problems and fixes
- "Your browser could not read audio from this file." The file is in a format your browser can't open, or it's damaged. Convert it to MP3 or M4A with any free audio converter and try again.
- The text is in the wrong language. You left Auto-detect on and it guessed wrong. Choose the language and transcribe again.
- Whole sentences are missing. Usually music or noise over speech. Try Accurate, or clean up the audio first.
- The transcript repeats a phrase. This can happen over long silences or music. Delete the repeats; they don't affect the rest.
- It's slow on a phone. Phones have less memory and power. Use a computer for long recordings, or split the file.
AI tools that turn transcripts into summaries and posts are coming soon to Spokly; transcription stays free.
FAQ
Can I convert an MP3 to text for free?
Yes. Spokly's audio and video to text tool transcribes MP3 files for free in your browser, with no sign-up and no limit on the number of files.
Is my audio uploaded?
Not with on-device transcription. The speech model runs in your browser, so the recording stays on your computer or phone.
How long can the MP3 be?
Up to 60 minutes per file on a computer and 15 minutes on a phone. Split longer recordings into parts.
Which languages are supported?
33 languages, including Arabic (with several dialects), English, French, Spanish, German, Turkish, Hindi, Urdu and Indonesian.
Will it tell me who is speaking?
No. The transcript is the words in order, without speaker names. Add names yourself as you read it through.