MP3 to text on iPhone and iPad, processed on the device

An MP3 already sitting on your phone does not need to travel to a server to become text. LoroNote imports the file, runs Whisper Large V3 Turbo locally on your iPhone or iPad, and exports the result as TXT or as an SRT with timecodes and speaker labels. No account, no minute meter, no upload.

Input
.mp3 at any bitrate
Output
TXT, or SRT with timecodes
Where it runs
On your iPhone or iPad
Internet
Not required
Price
Free download · $39.99 lifetime

How to convert an MP3 to text

  1. 1

    Install LoroNote and open it

    The download is free and no account is created — there is no sign-up screen, because there is no server holding your transcripts.

  2. 2

    Tap Import File

    Pick the .mp3 from Files, iCloud Drive, or a download folder. If the file arrived as an email attachment or a chat download, save it to Files first so the picker can see it.

  3. 3

    Confirm the language

    LoroNote detects the spoken language automatically and supports over 100. If the recording is full of names, product codes, or jargon, add them to the custom dictionary first so they come back spelled correctly.

  4. 4

    Let it transcribe on the device

    Whisper Large V3 Turbo runs on the phone or iPad itself — airplane mode changes nothing. Keep the app open while a long file processes; it uses the device GPU and pausing it costs you time.

  5. 5

    Review the speakers, then export

    Review the automatically separated speakers and give them real names, then export. TXT gives you the text to paste into notes or a document; SRT carries the timecodes for quoting or captioning.

Where MP3 files come from

MP3 is the interchange format: it is what things get exported as when they need to be small enough to email, download, or fit on a memory card.

How long it takes

MP3 files are small, but the clock runs on the length of the recording — a 40 MB podcast episode and a 40 MB field recording take very different amounts of time.

On-device transcription time by recording length, measured on an iPhone 15 Pro
Recording lengthTime to transcribe
5 minutesabout 16 seconds
30 minutesabout 1.5 minutes
1 hourabout 3 minutes
3 hoursabout 10 minutes

Measured by LoroNote on an iPhone 15 Pro at roughly 19x real time. Newer devices transcribe faster; older supported models take longer.

What decides whether an MP3 transcribes well

Bitrate matters more than length

Handheld recorders often save at 32–64 kbps mono to stretch a memory card. That is clear enough to follow by ear and thin enough to cost you accuracy on names, numbers, and acronyms. If the original recording still exists in a better format, transcribe that instead.

Mono is fine; distance is not

Stereo buys you nothing for speech. How far the speaker sat from the microphone, and how much the room echoed, set the ceiling on what any model can recover.

An MP3 carries no speaker information

Nothing inside the file says who was talking. The labels on your transcript come from LoroNote’s on-device speaker identification, which you then rename — the file itself contributes nothing.

A downloaded podcast is still someone’s copyright

Transcribing an episode for your own notes or research is one thing; republishing the transcript is another. The app will not stop you, so the judgement is yours.

What you give up to a free online MP3 converter

The web tools that rank for "MP3 to text" all work the same way: you hand them a copy of the file. That is fine for a podcast episode and a genuine problem for a recorded client call, a candidate interview, or a conversation with a source — you are trusting a retention policy you have not read, on infrastructure you cannot inspect.

They also cap you, because each minute they transcribe costs them GPU time on a server. Free tiers stop at a file size, a length, or a monthly allowance. LoroNote has no per-minute cost to recover, so a paid plan sets no cap on minutes or files — the constraint is your device, not a billing model.

Common questions

Can I transcribe an MP3 without an internet connection?

Yes. Both import and transcription work offline, because Whisper Large V3 Turbo runs on the device. This is the practical test for whether an app is genuinely local: turn on airplane mode and see whether the transcript still appears.

What is the longest MP3 I can transcribe?

There is no plan cap on minutes or file count. A three-hour recording takes roughly ten minutes to process on an iPhone 15 Pro. The real limits are free storage, battery, and keeping the app open while it works.

Will a low-quality MP3 still produce a usable transcript?

Usually yes, with more mistakes on proper nouns. A 32 kbps mono recording from a pocket recorder is intelligible but thin, so names and numbers are where errors cluster. Adding those names to the custom dictionary before transcribing recovers most of them.

Can I get timestamps from an MP3 transcript?

Yes. The transcript is timestamped at sentence level in the app — tap a line to replay that exact moment — and SRT export writes those timecodes into a standard subtitle file.

Do I have to convert the MP3 to another format first?

No. LoroNote imports MP3, WAV, M4A, MP4, and MOV directly. Converting between lossy formats before transcribing only degrades the audio.

Related

Point it at the MP3 you have been putting off

LoroNote is a free download. Import one file, watch it transcribe with the network off, and decide from the result.

Download LoroNote