Skip to content
Voice model · Hexgrad

Kokoro when you sell in English and do not want to configure anything

This engine has no emotion to choose, no sliders, and it reads no tags. You pick a voice, you type the copy, you get a recording — and for an English advert that is often the whole of what you need. It will not read Polish and we are not pretending otherwise: we checked it with the same sentence as the other nine.

Rendered by this model

Turn the sound on and listen

–:–– / –:––

The script the engine was given

Our new collection is in the shop today. Cotton, linen, and the one colour that always sells out first. Take a look, order before two, and it ships the same day.

Read in English by am_michael, because Polish on this engine comes back as phonetic mush — our test sentence returned as "NASA NOAA Kolkja jest jezdostepna". The English read is clean word for word, which is what it sounds like used for what it is for.This was generated by the model this page is about, exactly the way the app runs it. It is not a photo shoot and not client work.
This engine hands back a WAV. The file above is transcoded to MP3 so the page is not several megabytes — the voice and the length of the read are untouched.

Specification

Built by
Hexgrad
Speaks Polish
no — reads Polish text phonetically
Ready-made voices
20
Steering the delivery
speed and nothing else
File from the engine
wav — re-encoded to mp3 for this page
Render time
about 20s
Cost
1 credit per 1000 characters, rounded up

What it is good for

Nothing to set, so nothing to get wrong
A voice and a speed — that is the list. Across twenty English product descriptions the lack of choice is a feature: every file sounds the same, because there is no field that could drift between them.
Twenty voices, British and American among them
The list is closed, so the engine rejects a typo in a name rather than quietly substituting another voice. Britain and America get their own read from the same list.
Quicker than every other English-only engine here
Of the four that do not speak Polish, this one comes back by far the soonest — Orpheus TTS can take nearly three minutes on the same sentence. Across a series of descriptions that is an hour.

When to pick something else

It will not read Polish, and that is not an accent problem — Polish words come out as syllables with no meaning. If the advert is in Polish, use MiniMax Speech 2.8 HD or ElevenLabs Multilingual v2 and do not try to work around it with phonetic spelling. Even in English the delivery here is flat: no emotion, no tags, no laugh. When you want a recording with character there is Chatterbox HD and its expression slider, and when you want a laugh mid-sentence, Orpheus TTS.

See every model

How to brief this model

The things this engine handles differently from the rest. The habits that hold for every model are collected separately, in the help centre.

  1. Do not put Polish text in here, and do not try phonetic spelling. The engine has no Polish sounds — what comes out is English pretending to be Polish words, not a Polish advert.
  2. Pick the voice from the list the field publishes. It is closed, so a wrong name returns an error and you do not pay for a read in the wrong voice.
  3. Match the accent to the market. The American and British voices sound clearly different and, selling into one of those markets, it is audible from the first sentence.
  4. Leave the speed at one unless the read has to fit a slot. It is the only field that changes anything here, so any departure from the default is heard.
  5. Keep the sentences short. The engine does not steer intonation, so a long multi-clause sentence is read straight through with no breath at the comma.
How to write a brief that works

Questions about this model

Try this model on your own product

Sign up without a card and get 40 credits. Switching model is one click, so you can send the same brief through several of them and keep the best frame.