Turn recordings into text,
on your own PC.

Any file Windows can play

Audio or video, it makes no difference: MP3, WAV, M4A, MP4, MKV. Point at the file, choose where the transcript goes, and let it run in the background while you work.

A recording is evidence, so nothing is ‘corrected’

This is the one place Lickety deliberately does less. None of your formatting rules run on a transcript, and your vocabulary and replacements are off by default. A vocabulary prompt biases the recogniser toward words you supplied, which is a reasonable trade on your own dictation and the wrong trade entirely on somebody else’s recording, where it could insert a term nobody said. You can switch them on, and anything applied is named in the transcript’s own header so the next reader knows what was done to it.

What it does remove

Only the annotations the recogniser invents over silence and music — the “[BLANK_AUDIO]” and “[MUSIC]” lines. Spoken words are left alone, including the ones somebody mumbled.

How long it takes

It runs on your processor, so it depends on your machine and the model you pick. A twenty minute recording on a mid-range PC with the base model takes around three minutes. You choose how many cores to give it, so it can run quietly in the background or take the machine and finish sooner.

Is there a per-minute charge?

No. Transcription is included with Lickety and runs on your own computer, so the hundredth hour costs exactly what the first one did.

Does it identify who is speaking?

No. There is no speaker separation, so a two-person interview comes back as one run of text.

Is the recording uploaded anywhere?

No. The file is read from your disk and the transcript is written back to your disk.

What formats does it take?

Anything Windows can play, which in practice means MP3, WAV, M4A, MP4 and MKV among others.