GUIDE

Yukkuri voice alternative: AquesTalk, licenses and natural voices

The Yukkuri voice is one of the most recognizable sounds of Japanese internet video: two round Touhou heads explaining a topic in a flat synthetic voice. Many creators love it, and many others search for a voice that is easier to listen to for a long video, or for clearer terms when the channel starts earning.

This guide explains what the Yukkuri voice is, what its license says in general terms, and how SuperClonica can voice Japanese scripts with natural voices, free to try online and offline on your own computer.

Browse voices Install the SuperClonica engine

What the Yukkuri voice is

Terms change, so check AQUEST’s current license before you monetize videos made with the Yukkuri voice.

  • The Yukkuri voice comes from AquesTalk, a speech engine by the Japanese company AQUEST.
  • It became famous through Yukkuri commentary videos with characters from the Touhou Project fan scene.
  • AquesTalk is free for personal non-commercial use; commercial use requires a paid license.
  • The Yukkuri MovieMaker 4 FAQ explains its license requirement and offers a Lite edition without AquesTalk.

Hear a natural Japanese voice free

Open the page of any voice, type a Japanese line of 30 to 100 characters, choose Japanese and press Voice it. In a few seconds you can compare a natural voice with the Yukkuri sound, free and with no account.

When to choose which voice

SuperClonica does not imitate the Yukkuri voice; it offers different, natural-sounding voices.

  • Keep the Yukkuri voice when the classic meme style is the point of the video.
  • Choose a natural voice for long explanations that viewers listen to for minutes.
  • The ten stock voices of Supertonic 3 are free under the OpenRAIL-M license, commercial use included.
  • Catalog voices and voices trained from your own recording give a channel its own sound.

Voice a whole script offline

For a whole video script, install the free engine with python -m pip install superclonica-cpu, save the Japanese text in script.txt and run superclonica --text-file script.txt --voice F3 --lang ja --output script.wav. Long texts are split at natural pauses and joined into one file, with no character limit, and everything runs offline. Write numbers and readings the way they should be spoken: the engine reads digits as numbers.

Two-speaker commentary

Yukkuri videos are usually conversations between two characters. Give each speaker a different voice, a lower and a higher one, voice their lines as separate files and place them in your video editor, or use a dialogue script that the engine joins into one file. Say that the voices are AI-generated where viewers would care.