Someone speaks.
You read along.
The words appear on your screen as they are said — and when
you read them out, the line you are on is marked, and it keeps up with
your voice.
What the reader sees
It's part of the
virtual machine's own
memory
that holds every object
you have made so far.
How it works
- Start a session. One button. You get a link.
- Send the link. Message it, mail it, or read the room name down
the phone — whatever is easiest. Nothing to install at the other
end.
- Talk. Your browser does the transcribing, so no audio is sent
to us and none is kept.
- They read. Words arrive as you say them. When the reader reads
aloud, three words behind are shaded, the word they are on is marked,
three ahead are faint, and what is finished fades away.
Why the marker stays put
Words are matched six at a time, not
one at a time. Any single word turns up all over a transcript and is no
evidence at all; six in a row are as good as unique, and one misheard word
in the middle is outvoted by the five around it. Only text actually on
screen can be matched — which is what stops the marker leaping into a
paragraph from ten minutes ago.
Pricing
Questions
- Which browsers?
- Speaking needs Chrome or Edge, on a laptop or an Android phone — they
are the ones with speech recognition built in. Reading works
anywhere.
- Is my voice recorded?
- No. The transcribing happens in the speaker's own browser; only the
text reaches us, and it is held in memory for the session and then
forgotten.
- Does the reader need an account?
- No. Opening the link is the whole setup.
- Is there anything to cancel?
- No — there is no subscription. You buy minutes when you want them, and
they sit in your account until you use them.
- Do minutes run out if I don't use them?
- The free ones do: they're per day, and they come back the next day.
Bought minutes don't expire.