Levi Ackerman TTS: make him say anything
Levi from Attack on Titan says little, and every word lands: short orders, dry remarks, calm replies in the middle of chaos. That makes him a favorite for fan edits, fan dubs and memes, and a perfect fit for a text-to-speech voice.
The SuperClonica catalog has a fan-made Levi voice for Supertonic 3, one of the most liked anime voices on the site. This guide shows what it sounds like, how to write lines in character and how to voice whole scenes offline.
What the Levi voice sounds like
It is not endorsed by the creators of the series or its voice actors.
- Low and restrained.
- Short orders and calm replies suit it better than long speeches.
- Recorded in Russian, so it keeps a Russian accent in English, Japanese and other languages.
- Every word of the voice is fan-made: it is not the voice of the Japanese or English voice actor.
Make him say a line free
On the voice page, type a line of 30 to 100 characters, choose the language and press Voice it. In a few seconds you can listen and download the recording for your edit, with no account.
Write lines in character
Write your own lines rather than quoting the series word for word: they sound fresher and keep the script out of it.
- Short sentences ending with a full stop: Levi does not shout.
- Orders in the imperative: clean this, move, stay here.
- Dry remarks: his famous dislike of dirt is good material.
- Few exclamation marks: calm delivery is the whole character.
Fan dubs and edits offline
For a fan dub or a long edit, buy the voice once, download its file and install the free SuperClonica engine: python -m pip install superclonica-cpu. Save the lines in a text file and voice them with one command: superclonica --text-file levi.txt --voice levi.json --lang en --output levi.wav. The engine splits long texts at natural pauses and runs offline on an ordinary processor, so a whole fan episode costs nothing extra.
Fan fun, not deception
Use the voice for fan edits, fan dubs, memes and personal projects within the SuperClonica terms. Do not present the audio as an official clip or as a real voice actor’s words.