Skip to content
Voice model · Resemble AI

Chatterbox HD when an English advert should sound like you, not like a voiceover

Of the ten engines here only this one will take a recording of a voice and read your copy in it. For a one-person shop that is the difference between an advert that sounds like an advert and one that sounds like the owner — and selling on Etsy or Amazon, that is the whole argument. It also has an expression slider that no other English-only engine here offers. It will not read Polish.

Rendered by this model

Turn the sound on and listen

–:–– / –:––

The script the engine was given

Right, the new collection is in — and it is better than the photographs. Cotton, linen, and one colour I would not sit on. Order before two and it goes out today.

Read in English by Aurora, with expression raised to 1.1 and the 48 kHz mode on — which is where the audible shifts in tone come from, and Kokoro has none of them. There is no Polish version because our Polish sentence came back so distorted that speech recognition guessed Swahili.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
Resemble AI
Speaks Polish
no — reads Polish text phonetically
Ready-made voices
9
Your own voice
yes — upload a recording and it speaks in that voice
Steering the delivery
sliders: stability and expressiveness
File from the engine
wav — re-encoded to mp3 for this page
Render time
about 25s
Cost
1 credit per 1000 characters, rounded up

What it is good for

It reads in the voice from your recording
You upload a sample and the engine speaks in it from then on. No other model on this list does that — the other nine offer ready-made voices only, so a one-person brand sounds like every other one there.
Expression as a slider, not a tag
One number raises the expressiveness of the whole read. Set it once and every subsequent advert sounds the same, unlike tag-based steering, which you have to remember to insert.
A 48 kHz mode for anything going under video
A read at the higher sample rate does not fall apart after a video platform compresses it. That is when to turn the field on, and only then — on a phone speaker there is no difference to hear.

When to pick something else

It will not read Polish, and not as a matter of accent: our Polish sentence came out as "Naza nawa kolechajis" and speech recognition took the recording for Swahili. For a Polish advert use MiniMax Speech 2.8 HD or ElevenLabs Multilingual v2. It is also one of the slower engines of the ten and has only nine ready-made voices, so if cloning your own voice is not the point, Kokoro does the same English read sooner and from a longer list.

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. To clone a voice, upload fifteen seconds or so of a clean recording made in one room. A sample with reverb or background gives you a voice that sounds like that sample, because the engine reproduces the acoustics too.
  2. Keep expression around 1.0 to 1.2. Past roughly 1.5 the read stops sounding animated and starts sounding unwell.
  3. Only turn the 48 kHz mode on for material going under video. For plain listening it raises nothing but the file size.
  4. Do not type Polish here, phonetic spelling included. The engine has no Polish sounds and that is not something spelling can work around.
  5. Set your cloned voice once and stay with one sample. A new source recording is a new timbre, and between two adverts you can hear that two different people are speaking.
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.