CONNECTING TO KAMITSUBAKI_CITY...

KAFU

A CeVIO AI musical isotope built from KAF's vocal data, designed as a creative partner that expands the possibilities of vocal synthesis for creators.

KAFU
中 日 EN
Edit source
Contents

Overview

KAFU is the first voice synthesis software in KAMITSUBAKI STUDIO’s “Musical Isotope” series, built from the vocal data of virtual singer KAF using AI technology to generate highly realistic singing voices.

Operating on the CeVIO AI platform, KAFU is positioned not merely as a singing synthesis tool but as a “creative partner” for creators — designed to expand the possibilities of vocal synthesis through UGC (User Generated Content), enabling more creators to express their music with near-human singing voices.

Role and Creative Position

The Musical Isotope Project’s philosophy is that AI technology should not replace human singing but serve as a partner for creators. KAFU inherits the “girlishness” and “sense of distance” from KAF’s voice while gaining a unique “non-human” quality inherent to合成 voices. This voice quality — situated between真人 and virtual — has become an aesthetic resource within the creator community.

In the UGC ecosystem, KAFU has been widely adopted by Vocaloid and CeVIO creators, generating a vast body of derivative works. She is both an extension of KAF’s voice and an independent creative symbol.

Activity History

  • 2021: Released as the first product in the Musical Isotope series, concurrent with KAF’s joining of V.W.P.
  • Became one of the most popular voice banks on the CeVIO AI platform.
  • Participated in KAMITSUBAKI FES’24 and FES’25.
  • Widely used by creators in original songs, covers, and cross-media projects.

KAFU does not release traditional music作品 but exists as a tool within the creator ecosystem. Representative songs created using KAFU include:

  • KAF × KAFU “Ryuusenkei Mayday” (composed by HiFi-P)
  • A vast body of derivative works by UGC creators

Useful related entries:

Related Projects / Setting

KAFU is a core product of the “Musical Isotope Project,” initiated by KAMITSUBAKI STUDIO to develop voice synthesis software based on the vocal data of its virtual singers using AI technology. Following KAFU, SEKAI (based on ISEKAIJOUCHO) also joined the series.

Contact: [email protected]

References