Thank you for being here. By landing on this page, you are cordially invited to explore and try the new AI voicebanks. We appreciate your interest in our AI vocal synthesis project, and we're excited to have you as part of this journey. Your presence here signifies your curiosity and openness to experience the possibilities of AI-generated voices. We sincerely thank you for your participation and look forward to your engagement with these innovative voicebanks.
Exploring our vocal-synth characters is now simpler than ever!
We've also added more assets for our characters.
Printto Magicbeat is originally an UTAU since 2014. He is known for his ability to sing in many languages. He can sing up to 7+ languages in UTAU. His first DeepVocal voice library is recorded in Thai. Later, his Japanese DeepVocal library got released in 2019.
DiffSinger
UTAU
DeepVocal
Millefeuille
Tanya Magicbeat is originally a DeepVocal since 2020. Her voicer's main language is Thai but her DeepVocal voicebank supports Japanese. Later in 2024, she was upgraded to DiffSinger voicebank for OpenUTAU. Her DiffSinger voicebank supports Thai, English, and Japanese with the main language is Thai.
DiffSinger DeepVocal
You can also show your support for our partners!
Managed by Tempoprimo
Maguro San is a legendary tuna who loves listening to the music of sailors. Dreaming of becoming a singer like a human, Maguro San made a deal with a sea witch, trading all its powers—such as an extended lifespan and the ability to command a giant school of tuna—in exchange for a human-like form to pursue its dream.
DiffSinger UTAU Thai English Japanese Korean
Managed by Masahiro08_28
Altough open for interpretation, his personality is often depicted as a kind of tranquil, well meaning and a bit aloof but kind person. He is often portrayed to be friends with Sayo and is the older brother of Yuri Kone. His character birthday is 08/28.
DiffSinger UTAU Japanese English Thai French Chinese Spanish Korean
Managed by MahrFah_
Zephyr hails from a dimension where music and performance hold unparalleled power, capable of shaping reality and bridging worlds. He was chosen as an ambassador to share his realm's culture with others, spreading harmony (and a bit of chaos) through his captivating performances. While he thrives in the spotlight, he also carries the weight of representing his people, balancing his carefree demeanor with a sense of responsibility.
DiffSinger Thai English Japanese
Managed by Toshio【年男】
Toshio is a middle-aged wolf who works as a singer. He has Dissociative Identity Disorder, which means he has multiple identities within one body, allowing him to sing in various languages depending on the personality that takes control at any given moment. He often sleepwalks and frequently wanders absentmindedly, sometimes crashing into things due to his confusion and dazed state, unsure of which identity or personality is currently in control.
DiffSinger UTAU Thai English Japanese French Chinese Spanish
Managed by paperminttyy
Mairu Maishi is a name crafted with playful wordplay—meant to be read in a Japanese style as "ma-i-ru ma-i-shi". Its sound hints at the Thai phrase "ไม่รู้ ไม่ชี้", but her origins remain unknown—just like her story.
DiffSinger UTAU Thai English Japanese French Chinese Spanish Korean
Managed by ROO310
Lucy Roos or Nasuka Roos has been an UTAU since 2016. It is a furloid from Thailand. In 2018, Nasuka Roos stopped development due to lack of equipment and time, so it disappeared for 8 years. In 2025, Nasuka Roos brought it back to develop again in the DiffSinger project.
DiffSinger UTAU Thai English Japanese
Our partnered voice libraries utilize these publicly available singing datasets for each languages:
Thai: Printto TH
English: TIGER
, AlexFloarea
, and Project AI❤️dol
Japanese: Ofuton P
and PJS
(relabeled by Utautautau)
French: PetitMillefeuille
Korean: CSD
(relabeled by heta-tan)
Chinese: M4Singer
(relabeled by nobodyP)
Spanish: Namine Criss