How is this different from the HD tier?

In the wait and in the quality of the recording. The controls, voices, emotions and character limit are identical, and both cost the same here — which is why Turbo makes sense for auditioning and HD for the take that goes into the advert.

Why did it get "330 ml" wrong?

Because it expands shorthand its own way and there is no field to tell it otherwise. The fix is on your side: write "trzysta trzydzieści mililitrów". The engine that does it for you is ElevenLabs Multilingual v2, which read "24 h" back to us as "24 godziny".

What does a read cost?

One credit per 1000 characters started. An advert script is usually around five hundred characters, so one read is one credit.

Voice model · MiniMax

MiniMax Speech 2.8 Turbo when you are listening to five versions of the same line

This tier exists so you are not waiting. The voices, the emotions, the pause markers and the 10,000-character limit are identical to HD — the difference is in the quality of the recording and in how quickly it comes back. While you are still deciding whether a line should be happy or neutral, audition it here and record the take you actually publish on HD.

Claim 40 credits See pricing

Rendered by this model

Turn the sound on and listen

–:–– / –:––

The script the engine was given

Ten kubek ma 330 ml, grube ucho i szkliwo, które nie matowieje w zmywarce. Pakujemy go w tekturę bez folii, więc wyślesz go dalej w tym samym kartonie. Zamówienia złożone do 14:00 wychodzą tego samego dnia.

Read in Polish by Deep_Voice_Man. The numbers were deliberately left short in the script and this is exactly where the engine trips: "330 ml" came back as "330 milinów" and "do 14:00" as "do 14 zero". The rest of the paragraph is word for word.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.

Specification

Built by: MiniMax
Speaks Polish: yes — checked on a Polish sentence
Ready-made voices: 17
Steering the delivery: an emotion picked from a list, as a setting
Character limit: 10,000
File from the engine: mp3
Render time: about 8s
Cost: 1 credit per 1000 characters, rounded up

What it is good for

The shortest wait in the MiniMax pair: A short line comes back in seconds, so five versions of a headline are auditioned faster than you can read them out loud. While you are still finding the wording, that matters more than the last few percent of quality.
Nothing to rewrite when you switch: The same seventeen voices, the same seven emotions, the same pause markers and the same 10,000-character limit. A script tuned here goes into HD unchanged — you switch the engine and record it again.
A whole product description in one job: Ten thousand characters is several pages. Hearing all of it as one read shows you where the copy drags, which reading it paragraph by paragraph on screen does not.

When to pick something else

It costs the same credit as HD, so the only thing you are buying is the wait — record anything you publish again on MiniMax Speech 2.8 HD. Separately, it reads shorthand its own way: "330 ml" came out as "330 milinów" and "do 14:00" as "do 14 zero", which you can hear in the sample. If your copy is full of clock times and units and you would rather not expand them by hand, ElevenLabs Multilingual v2 does it itself. And if all you want is the shortest wait of the ten, that is ElevenLabs Turbo v2.5.

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.

Turn shorthand into words before you send the script. This engine gave us "330 milinów" for "330 ml" and "14 zero" for "14:00" — the one repeatable fault it has on Polish copy.
Keep this tier for auditioning, not for the final take. The same words sound better on HD for the same credit and the script carries over unchanged.
While you are hunting for a delivery, change the emotion rather than the words. The same advert in happy and in neutral is two different recordings and each costs the same.
Do not try to mix voices in one file here — this engine reads with one. A two-person dialogue is what VibeVoice 7B is for; it takes up to four voices in one script.
Set the language field to Polish. Without it you get Polish read with an English accent, and that is the field rather than the voice.

How to write a brief that works

Questions about this model

Other models

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.

Claim 40 credits Work out your saving