Canary (カナリー ) is a Japanese voicebank from JAE VOCAL PROJECT that was developed and voiced by Mina.Ki. He is the first character from JAE VOCAL PROJECT to be released for DeepVocal.


Canary was originally released for the UTAU engine on February 2 and was the third character of the series overall. He was known as JAE-03.[1] On June 21, his Japanese voicebank was released.[2]

Canary was expected to receive DeepVocal​ voicebanks along with UTAU.[3]​ He would be the first in the JAE VOCAL PROJECT to be available for the software. His Japanese DeepVocal voicebank was released on August 8, 2019.[4]​ The illustrator was confirmed to be Oweruru.[5]


Many years ago on a fateful day, Canary randomly arrived on Earth. He spends his days hanging out with robots and his boyfriend who happens to be a zombie.

Canary likes food, his boyfriend, singing, animals, plush toys, and drawing​. He dislikes pretentious people​.


Canary was named after the songbird. It is also the name of a shade of yellow.​

His name in Kana is rendered as "カナリー".


Canary has blond hair tied in a top bun. His eyes are yellow with a green tint at the bottom.

On his face are yellow lip piercings and a heart tattoo under his left eye. He wears a winter outfit with a red and periwinkle gradient, a transparent red scarf, and pants with boots. Overall, he has a feminine appearance.


Product information



嘘とぬいぐるみ (Uso to Nuigurumi) - Short version Niconico YouTube bilibili SoundCloud
メガネの向こうの世界 (Megane No Mukou No Sekai) Niconico YouTube bilibili SoundCloud

Voicebank libraries

Product Information
  Genre: TBA  Tempo: TBA  Range: TBA
Basic Information
Details are currently unknown, but the Japanese voicebank uses the same recordings as the UTAU counterpart. This voicebank is described to be an energetic vocal with a sharp tone. It uses Zurui's reclist for CVVCV.
Bank/Phonetic Information
  • Japanese
    • Issues: His hiragana dictionary is inaccurate and not correct for certain characters
Voicebank Sample

Canary - Japanese




External links


DeepVocal Voices

Community content is available under CC-BY-SA unless otherwise noted.