Vocaloid
 | This article is a work in progress. ...Well, all the articles here are, in a way. But this one moreso, and the article may contain incomplete information and editor's notes. Notes: culture |
Vocaloid is a vocal synthesis software, internet culture, and way of life. There are many different kinds of Vocaloid, from the well-known alternative UTAU, to the pop producer and famed Hatsune Miku collaborator DECO*27, to whatever the fuck producers like ぬぬぬぬ are up to. The program can be used to input notes and phonemes to have a selected voicebank sing a pattern on playback. There are many similar softwares, either competing, or similarly in speech synthesis (text-to-speech).
Vocal synth
Vocaloid is the most famous, and one of the oldest, with VOCALOID 1 dating to 2004, the first two voicebanks being the pair LEON and LOLA. Their cover images are just stock images with text and a filter. The software is by far the most influential, hosting the likes of Hatsune Miku, Megurine Luka, Megpoid, KAITO, and many more. The first generation had not much of note, except for the two other voicebanks by Crypton Future Media, MEIKO and KAITO. VOCALOID 2 would be where everything started spicing up, with Miku and the rest. Other less renowned characters had also shown up around this time, like AH-Soft's Kaai Yuki, SF-A2 miki, and that other guy. Voices can be tuned with a set of parameters, shaping the character of the sound in various ways; DECO*27 has a very distinct tuning of Hatsune Miku, for example. Many popular characters have Append voicebanks, featuring a different vocal tone.
From there on, everything was set in motion, and Vocaloid continued receiving major versions about every four years.
UTAU is a free program similar to Vocaloid, albeit being free and therefore... Look, all I'm saying is when you have one of the most popular girl on the internet, you can afford some quality programming.
One of the big features is that voicebank creation and editing is included, allowing anyone with a microphone and free time to make their own. UTAU brings us the likes of Adachi Rei, Kasane Teto, Utane Uta (Defoko), and Momone Momo, not-very-well-known to be the voice of Nyan Cat[1]. (Yes, daniwell made the original[2], which featured Miku, but the original video with the Nyan Cat featured a cover with Momone Momo on it[3].) A large majority of all the total voicebanks ever are UTAU, due to the accessibility of creation. It still takes a lot of time and effort, but that's free in comparison to the unknown vast sums it would cost for any other engine
Despite appearing as a distinct character, Akita Neru does not have her own official voice, usually being derivative of Hatsune Miku or Kagamine Rin[4]. Other derivative characters include Yowane Haku.
CeVIO Creative Studio.[was] It won an award from Microsoft, apparently. Speech synthesis is available for supported voicebanks.
Synthesizer V was originally released in 2018, later dubbed Synthesizer V R1 for clarity[5]. Deriving from an advanced library for an alternate resampler for UTAU[6][7], the first generation wasn't much of note, with the only name having much of any fame being Eleanor Forte. 2020 brought the release of Synthesizer V Studio, being among the first to adopt neural network based synthesis and helped bring 'AI' to everything vocal synth. Synthesizer V Studio and AI voicebanks gave impressive fidelity, and while lacking the fine-tune control of Vocaloid tuning, it brings Vocal Modes allowing for a much wider range of sounds from the same voicebank, doing away with the need for Appends. Synthesizer V Studio 2 released in 2025, bringing generational uplifts, reducing engine noise and enhancing tuning controls.
Synthesizer V Studio, Vocaloid 6, and ACE Studio support multilingual synthesis, opening up voicebanks on those platforms to sing perfectly in several languages, mostly Japanese, Chinese, and English.
NEUTRINO, also stylized as n3utrino, is described as a "Neural singing synthesizer"[8], and initially released in 2020 effectively as an early demonstration of AI voicebanks. It kind of just... exists, but at least it's more notable than CeVIO Creative Studio. A few well-known names showed up here, before mostly moving on to more renowned software: Kotonoha Aoi and Akane, Tohoku Zunko and... oh, it's just first-party voices, the Kotonohas, and various SSS LLC characters.
CeVIO AI.[is] It has voices and characters, including noteworthy ones like KAFU, IA, Ci flower, and has a very present engine noise shaping the sound.
VoiSona is CeVIO AI but more is. And also subscription service model. It has voices and characters, including noteworthy ones like Chis-A,
New Type (Piapro Studio NT) is a singing synthesizer that only has Piapro characters. Well, maybe only Hatsune Miku and Kagamine Rin/Len. Okay, basically only Miku.
ACE虚拟歌姬 (ACE Studio)... Well, it's also there, but it's moreso just there in the back rather than the distinct void of existence that is CeVIO. It's Chinese, subscription service, and supports creating voicebanks, similar to UTAU. It has Luo Tianyi, one of the original Vsingers.
Text-to-speech
Text-to-speech has an extensive history largely unrelated to Vocaloid, but famous or relevant ones include DECtalk, being the voice of Stephen Hawking and the game Moonbase Alpha, and AquesTalk, whose Female Voice 1 is the provider for Utane Uta.
The majority of its scene influence is on the Japanese side of the internet, primarily Niconico and part of YouTube, with many of the characters here having much more clearly defined traits, characters, and personalities, either via consensus or specific authors' choices. By sheer volume, (i'm sure) a substantial amount of all vocal synth consists of voiceovers, let's-plays, and anything that could be done with having full access to a voiced character.
VOICEROID is the first speech synthesis offering from AH-Soft, originally releasing in late 2009 using the AITalk speech synthesis engine. At first, not much was going on, but VOICEROID+, out a year later, soon brought many more iconic characters. VOICEROID+ EX was an upgrade from that in 2014, followed by VOICEROID2 in 2017. This and similar later software supports inputting a script for one or more characters, to optionally construct dialogue, with adjustment options depending on specific program. Other similar but independent character-specific products were built on the same AITalk engine, including Gynoid TALK, Talk Ex, and galaco Talk[9], including a limited set or just one character each. The engine noise is clearly defined breaks between samples, similar to Vocaloid.
Characters include Tsurumaki Maki, Yuzuki Yukari, Kizuna Akari, Kotonoha Aoi, Kotonoha Akane, Iori Yuzuru, Tohoku Zunko, Tohoku Kiritan, and others.
A.I.VOICE is effectively the sequel to VOICEROID2. Like Voiceroid, the primary language is Japanese, but now a few supporting voicebanks are available for English and Chinese. Most characters from VOICEROID have versions here, along with many more additional ones, including Adachi Rei and flower. Despite its recent release, it doesn't seem to use AI voicebanks.
VOICEVOX is a fully freely available[10] speech and vocal synthesis software utilising neural network technology. It features many defining characters, including Zundamon, Kasukabe Tsumugi, Tohoku Kiritan, Nurse Robot_Type T, and other related ones. Like VOICEROID, the program and documentation are only in Japanese. The engine noise is that of typical neural networks, being a fluid and coherent sound with marginally reduced audio fidelity.
VOICEPEAK is a speech synthesis software by AHS and Dreamtonics, the developers of Synthesizer V. The main features are the AHS characters and very high fidelity synthesis. It features characters such as Kasane Teto, Tsurumaki Maki, Tohoku Kiritan, Haruno Sora, and Hanakuma Chifuyu.
CoeFont and COEIROINK are two unrelated but similar text-to-speech with AI tech. Both support creating voicebanks, and have free options, with CoeFont having a paid version. COEIROINK doesn't include notable characters, but CoeFont has SAYO, Kazehiki, Nurse Robot_Type T, Yokune Ruko, Allial, Millial, and Averuni. The latter three are first-party and multilingual.
List of characters
Crypton Future Media (Piapro)
| Character | Programs | Notes |
|---|---|---|
| MEIKO | Vocaloid (1, 3, 4) | All Piapro characters have multiple Append voicebanks. |
| KAITO | Vocaloid (1, 3, 4) | |
| Hatsune Miku | Vocaloid (2, 3, 4, 6), New Type | 初音ミク. New Type[is] |
| Kagamine Rin | Vocaloid (2, 4), New Type | |
| Kagamine Len | Vocaloid (2, 4), New Type | |
| Megurine Luka | Vocaloid (2, 4) |
VOCALOMAKETS
| Character | Programs | Notes |
|---|---|---|
| Yuzuki Yukari | Vocaloid (3, 4, 6), Voiceroid (+, +EX, 2), A.I.VOICE (1, 2), Seiren Voice | Has multiple variants, Appends, and unique costumes for most. |
| Kizuna Akari | Vocaloid (4, 6), Voiceroid (2), A.I.VOICE (1, 2), Voidol, Seiren Voice | Has multiple variants. Voidol and Seiren Voice are voice changers. |
| Yuzuki Yukari (Shizuku) | A.I.VOICE (1) | A younger derivative of Yuzuki Yukari, and a distinct character. |
| Kizuna Akari (Tsubomi) | A.I.VOICE (1) | A younger derivative of Kizuna Akari. |
| Kizuna Akari (Moe) | A.I.VOICE (1) | An even younger derivative of Kizuna Akari. |
AHS (AH-Soft)
A few nobody gives a fuck about may have been omitted. Feel free to add those back if someone actually does care.
| Character | Programs | Notes |
|---|---|---|
| SF-A2 開発コード miki | Vocaloid (2, 4), Synthesizer V (Studio 2) | SF-A2 codename miki. Voiced by and named after Furukawa Miki |
| Kaai Yuki | Vocaloid (2, 4) | Voiced by a child. Famous for Lagtrain. |
| Hiyama Kiyoteru | Vocaloid (2, 4), Synthesizer V (Studio 2) | The teacher of Kaai Yuki. All 3 released on the same day. |
| Tsurumaki Maki | Voiceroid (+, +EX), CeVIO AI, VoiSona, Synthesizer V (Studio 1, 2), VOICEPEAK | Voiceroid has a different voice actor. Has 5 total Synth V voicebanks. |
| Kotonoha Aoi | Voiceroid (+, 2), A.I.VOICE (1, 2), Voidol, NEUTRINO, Synthesizer V (Studio 1), Seiren Voice, Vocaloid (6) | Seiren Voice and Synthesizer V have one voicebank for both sisters. |
| Kotonoha Akane | Voiceroid (+, 2), A.I.VOICE (1, 2), Voidol, NEUTRINO, Synthesizer V (Studio 1), Seiren Voice, Vocaloid (6) | Both share a voice provider, but Akane has a Kansai dialect. |
| Nekomura Iroha | Vocaloid (2, 4), Synthesizer V (Studio 2) | Themed after Hello Kitty. Has a vast vocal range. Voiced by a trans guy |
| Minase Kou | Voiceroid (+EX), VOICEPEAK | |
| Kyomachi Seika | Voiceroid (+EX), Synthesizer V (Studio 1, 2), VOICEPEAK | Her voice dispels any focus. |
| Tsuina-chan | Voiceroid (2), Synthesizer V (Studio 1, 2), Vocaloid (6) | |
| Miyamai Moca | Synthesizer V (Studio 1, 2), VOICEPEAK | Draws the focus Seika dispels. |
| Haruno Sora | Vocaloid (5), Voiceroid (2), Synthesizer V (Studio 1, 2), VOICEPEAK | Contains whimsy and sillyness. Has a deep and soft voice. |
| Iori Yuzuru | Voiceroid (2), A.I.VOICE (1,2), Seiren Voice | |
| Frimomen | Synthesizer V (Studio 1, 2), VOICEPEAK | Included with VOICEPEAK voicebanks. Has a superhero transformation. |
| Super Frimotan | VOICEPEAK | Girl Frimomen. Available with 5 Frimomen activation codes. |
| Asumi Ririse | VOICEPEAK, Synthesizer V (Studio 2) | |
| Asumi Shuo | VOICEPEAK, Synthesizer V (Studio 2) |
INTERNET Co., Ltd
As AHS has moved over to Synthesizer V, Internet Co. handles many Vocaloid 6 ports of voicebanks. A few nobody gives a fuck about may have been omitted.
| Character | Programs | Notes |
|---|---|---|
| GUMI | Vocaloid (2, 3, 4, 6), Voidol, A.I.VOICE (1, 2), Synthesizer V (Studio 1, 2), own talk software | The character is GUMI, but the product name is Megpoid (or AI Megpoid). |
| galaco | Vocaloid (3, 6), own talk software, A.I.VOICE (1, 2) | In lowercase. First split into a Red and Blue duo, with White and Black for V6. |
| Otomachi Una | Vocaloid (4, 6), own talk software, Voiceroid (2), Voidol, VOICEPEAK, Synthesizer V (Studio 1, 2), A.I.VOICE (2) | On 4 separate talk softwares. AHS do Synthesizer V and VOICEPEAK. |
| Kamui Gakupo | Vocaloid (2, 3, 4) | |
| Hibiki Koto | Vocaloid (6), Synthesizer V (Studio 1, 2) | [is] |
| ROSA | CeVIO AI, Synthesizer V (Studio 1) | Sister of CUL. |
| CUL | Vocaloid (3) | Unfortunate name. |
| kokone | Vocaloid (3) | [was][is] |
| Chika | Vocaloid (3) |
KAMITSUBAKI STUDIO
The label consists of 'Virtual Isotope Phenomenon'[11], the vocal synth characters, and 'Virtual Witch Phenomenon'[12], the corresponding VTubers.
| Character | Programs | Notes |
|---|---|---|
| KAFU | CeVIO AI | Voice from VTuber KAF. A Synthesizer V AI voicebank was planned, but was postponed in 2024 and canceled in 2025. |
| SEKAI | CeVIO AI, VOICEPEAK | Voice from VTuber Isekaijoucho. |
| RIME | CeVIO AI, VOICEPEAK | Voice from VTuber RIM. |
| COKO | CeVIO AI, VOICEPEAK | Voice from VTuber KOKO. |
| HARU | CeVIO AI, VOICEPEAK | Voice from VTuber Harusaruhi. |
TOKYO6 ENTERTAINMENT
Voicebanks are distributed by AHS.
| Character | Programs | Notes |
|---|---|---|
| Koharu Rikka | Synthesizer V (Studio 1), CeVIO AI (talk voice), VOICEPEAK | The trio attend the same high school.[13] Has Synth V Standard and AI voicebanks.
shes great backing vocals --ChifuyuDownTheLine (talk) 03:50, 3 December 2025 (UTC) |
| Hanakuma Chifuyu | Synthesizer V (Studio 1), CeVIO AI (talk voice), VOICEPEAK | my favorite --ChifuyuDownTheLine (talk) 03:50, 3 December 2025 (UTC) |
| Natsuki Karin | Synthesizer V (Studio 1), CeVIO AI (talk voice), VOICEPEAK | All 3 are confirmed to receive a Synth V upgrade in February 2026.
Most known for being on SICK -やんでるEP- by ゆよゆっぺ, the track 'Datte' having a few million YouTube views. |
SSS LLC. (Zunko)
| Character | Programs | Notes |
|---|---|---|
| Tohoku Zunko | Voiceroid (+, +EX), Vocaloid (3, 4), Voidol, NEUTRINO, CeVIO AI, VoiSona, VOICEPEAK, VOICEVOX | She has a wikipedia page for her family. |
| Tohoku Kiritan | UTAU, Voiceroid (+EX), Voidol, NEUTRINO, CeVIO AI, VoiSona, VOICEPEAK, Seiren Voice, VOICEVOX | Named after the region. Surname either Tohoku, Touhoku, or Tōhoku. |
| Tohoku Itako | UTAU, Voiceroid (2), NEUTRINO, CeVIO AI, VoiSona, Seiren Voice, VOICEPEAK, VOICEVOX | The oldest of the three. |
| Zundamon | UTAU, VOICEVOX, NEUTRINO, Seiren Voice, VOICEPEAK, CeVIO AI, VoiSona, Paravo (Parakeet.VC),
Voiceger (GPT-SoVITS)[only result for platform] |
The creature-est. Themed after zunda-mochi. Platform testbed. |
| Shikoku Metan | UTAU, VOICEVOX, NEUTRINO, CeVIO AI, VoiSona | All characters are part of the anime musical 'Zunda Horizon'. |
| Kyuushuu Sora | UTAU, VOICEVOX | |
| Chuugoku Usagi | UTAU, VOICEVOX | |
| Ooedo Chanko | UTAU, NEUTRINO, VOICEPEAK | |
| Chuubu Tsurugi | UTAU, VOICEVOX | |
| Kansai Shinobi | UTAU | |
| Hokkaido Meron | UTAU | Official website illustration shows she does have at least one melon. |
| Okinawa Awamo | UTAU |
First-party (Yamaha, Dreamtonics)
A majority of Yamaha's voicebanks are either developed for other groups, or are talking heads optionally in silhouette. Few have defined characters. Dreamtonics' voicebanks have minimal cover art.
| Character | Programs | Manager | Notes |
|---|---|---|---|
| asa | Vocaloid (Mobile) | Yamaha | Mystery! |
| Po-uta | Vocaloid (6) | Yamaha | Porter Robinson. Has strict terms. |
| Nurse Robot_Type T | UTAU, CoeFont, COEIROINK, VOICEVOX, Vocaloid (6), VoiSona | Yamaha | Originally an UTAUloid. |
| Mai | Synthesizer V (Studio 1, 2) | Dreamtonics | Included with Synthesizer V Studio. |
| Eleanor Forte | Synthesizer V (R1, Studio 1, 2 "Plus") | Dreamtonics | ENG-F1. Has a 'Plus' voicebank trained on the Standard files. |
| Hǎiyī | UTAU, Synthesizer V (R1, Studio 1, 2 "Plus") | Dreamtonics | Private UTAU voicebank, derived to Synth V R1. Fun! |
| Yì Xī | Synthesizer V (Studio 1, 2) | Dreamtonics | Meat girl. ^q^ |
Gynoid
| Character | Programs | Notes |
|---|---|---|
| flower | Vocaloid (3, 4), own talk software, CeVIO AI, VoiSona, A.I.VOICE | Product as v flower or Ci flower. Ci flower is different, but still flower. |
| Meika Hime | Vocaloid (5), own talk software, A.I.VOICE (2, 1) | unspecified gender. Maybe enby! Maybe vagueposting. Sibling of Mikoto. |
| Meika Mikoto | Vocaloid (5), own talk software, A.I.VOICE (2, 1) | unspecified gender. Maybe enby! Maybe vagueposting. Sibling of Hime. |
| Xin Hua | Vocaloid (3, 4) |
Eclipsed Sounds
| Character | Programs | Manager | Notes |
|---|---|---|---|
| SOLARIA | Synthesizer V (Studio 1, 2) | Eclipsed Sounds | Themed after the sun. |
| ASTERIAN | Synthesizer V (Studio 1, 2) | Eclipsed Sounds | Themed after the moon. |
| SAROS | Synthesizer V (Studio 1, 2) | Eclipsed Sounds | Nonbinary. Themed after stars. |
| NYL | Synthesizer V (Studio 1, 2) | Eclipsed Sounds | Nonbinary. The original 4 are incredibly stylish. |
| HXVOC | Synthesizer V (Studio 2) | Eclipsed Sounds | Mechanical arms. |
| GALENAIA | Synthesizer V (Studio 2) | Eclipsed Sounds | Has a very nice logo. |
Other groups
| Character | Programs | Manager | Notes |
|---|---|---|---|
| Kasane Teto | UTAU, Renoid, TALQu, Synthesizer V (Studio 1, 2), VOICEPEAK | TWINDRILL | Synthesizer V and VOICEPEAK distributed by AHS. |
| AVANNA | Vocaloid (3) | Zero-G Ltd. | One of the most famous English-only characters. Appears on two Porter Robinson songs. |
| Chis-A | VoiSona | Techno-Speech Ltd. | Main character of VoiSona. |
| Adachi Rei | UTAU, A.I.VOICE (1), DiffSinger | Mechanical Girl LLC. | Robot girl with a synthetic voice. Exists in real life. May receive a VOCALOID5 version. |
| Tsukomo Shion | UTAU | (missile_39) | Synthesis by the same author as Adachi Rei. Has a thousand-mile stare. |
| IA | Vocaloid (3, 6), CeVIO Creative Studio, CeVIO AI, VoiSona | 1st PLACE (IA PROJECT) | The first ARIA sister. Stylized as IA -ARIA ON THE PLANETES- |
| OИE | CeVIO Creative Studio, CeVIO AI, VoiSona | 1st PLACE (IA PROJECT) | The other ARIA sister. Stlyized as OИE -ARIA ON THE PLANETES- |
| Aoki Lapis | Vocaloid (3), CoeAvatar | i-style Project | CoeAvatar is unrelated to both CoeFont and COEIROINK. She has NFTs |
| LUMi | Vocaloid (4) | Akatsuki Virtual Artists | Included for a page link. |
| Fukase | Vocaloid (4) | TOKYO FANTASY | Voiced by Fukase Satoshi of the band managed by that company. |
Trivia
- Most of the program names listed are uppercase. This might be related to some Japanese input methods being based on western QWERTY layouts, and transcribing lowercase inputs to kana or kanji as typed/selected, but keeping capital letters as they are.[entirely vibes-based and anecdotal]
- Despite the common mentions of AI, generated imagery or backing instrumentals are not looked kindly upon. In voicebanks, it is purely a technological advancement, with usually lower storage cost and higher fidelity at the cost of compute.
References
- ↑ Nyan Cat - saraj00n feat. 桃音モモ (Music PV), from VocaDB
- ↑ Nyanyanyanyanyanyanya! - daniwell feat. 初音ミク (Original song), from VocaDB
- ↑ Nyanyanyanyanyanyanya! - ももももP feat. 桃音モモ (Cover), from VocaDB
- ↑ 亞北ネル, from VocaDB
- ↑ Synthesizer V R1, from VocaDB
- ↑ Moresampler, from VocaDB
- ↑ Moresampler - Kanru Hua's Website, archived on Wayback Machine
- ↑ NEUTRINO, from VocaDB
- ↑ AITalk, on VocaDB
- ↑ VOICEVOX on GitHub
- ↑ V.I.P, from VocaDB
- ↑ V.W.P, from UtaiteDB
- ↑ 小樽組, from VocaDB