voicevox

README

voicevox

English | 日本語

Go bindings for VOICEVOX core, for Japanese text to speech that runs locally.

• Uses purego and ffi, so CGo is not needed.
• Targets voicevox_core 0.17.0.

Installation

libffi

On Linux, install libffi.

sudo apt install libffi8

On macOS and Windows x64, the ffi package ships libffi, so you don't need to install it.

VOICEVOX runtime

Get the official downloader from the voicevox_core releases, then run it:

curl -sSfL https://github.com/VOICEVOX/voicevox_core/releases/download/0.17.0/download-linux-x64 -o download
chmod +x download
./download -o ~/voicevox_core

The downloader asks you to accept the VOICEVOX voice model terms and the VOICEVOX ONNX Runtime terms. To get only the talk models, add --models-pattern '[0-9]*.vvm'.

This creates the following layout:

voicevox_core/
  c_api/lib/libvoicevox_core.so
  onnxruntime/lib/libvoicevox_onnxruntime.so.<version>
  dict/open_jtalk_dic_utf_8-1.11/
  models/vvms/*.vvm

LoadOnnxruntime uses the recommended ONNX Runtime version if it is in the directory. Otherwise it uses the newest version it finds there.

Example

package main

import (
	"os"

	"github.com/hybridgroup/voicevox/pkg/voicevox"
)

func main() {
	voicevox.Load("voicevox_core/c_api/lib")

	ort, _ := voicevox.LoadOnnxruntime("voicevox_core/onnxruntime/lib")
	ojt, _ := voicevox.NewOpenJtalk("voicevox_core/dict/open_jtalk_dic_utf_8-1.11")

	syn, _ := voicevox.NewSynthesizer(ort, ojt, voicevox.DefaultInitializeOptions())
	ojt.Delete()
	defer syn.Delete()

	model, _ := voicevox.OpenVoiceModelFile("voicevox_core/models/vvms/0.vvm")
	syn.LoadVoiceModel(model, voicevox.DefaultLoadVoiceModelOptions())
	model.Delete()

	wav, _ := syn.TTS("こんにちは", 0, voicevox.DefaultTTSOptions())
	os.WriteFile("out.wav", wav, 0644)
}

The output is a 24 kHz, 16-bit, mono WAV file.

To change speed, pitch, intonation or volume:

1. Call CreateAudioQuery.
2. Edit speedScale, pitchScale, intonationScale or volumeScale in the returned JSON.
3. Pass the JSON to Synthesis.

A full example is in examples/hello:

go run ./examples/hello -data ~/voicevox_core -style 0 -text "こんにちは"

Voices

These are the talk styles in the 0.16.0 voice models from the downloader. Pass the id as the style to TTS or CreateAudioQuery, and load the model file that contains it. To list the styles at runtime, call VoiceModelFile.Metas() or Synthesizer.Metas(). The romanized names come from the character ids in the VOICEVOX website source.

Each character has its own terms. See the VOICEVOX website for details.

43 characters, 127 styles

Character	Styles (id, name)	Model files
四国めたん (Shikoku Metan)	2 ノーマル (normal), 0 あまあま (sweet), 4 セクシー (sexy), 6 ツンツン (prickly), 36 ささやき (whisper), 37 ヒソヒソ (hushed)	0.vvm, 5.vvm
ずんだもん (Zundamon)	3 ノーマル (normal), 1 あまあま (sweet), 5 セクシー (sexy), 7 ツンツン (prickly), 22 ささやき (whisper), 38 ヒソヒソ (hushed), 75 ヘロヘロ (exhausted), 76 なみだめ (teary)	0.vvm, 5.vvm, 15.vvm
春日部つむぎ (Kasukabe Tsumugi)	8 ノーマル (normal)	0.vvm
波音リツ (Namine Ritsu)	9 ノーマル (normal), 65 クイーン (queen)	3.vvm
雨晴はう (Amehare Hau)	10 ノーマル (normal)	0.vvm
玄野武宏 (Kurono Takehiro)	11 ノーマル (normal), 39 喜び (joy), 40 ツンギレ (snappy), 41 悲しみ (sadness)	4.vvm, 10.vvm
白上虎太郎 (Shirakami Kotarou)	12 ふつう (normal), 32 わーい (yay), 33 びくびく (nervous), 34 おこ (angry), 35 びえーん (crying)	9.vvm
青山龍星 (Aoyama Ryusei)	13 ノーマル (normal), 81 熱血 (passionate), 82 不機嫌 (grumpy), 83 喜び (joy), 84 しっとり (calm), 85 かなしみ (sadness), 86 囁き (whisper)	15.vvm
冥鳴ひまり (Meimei Himari)	14 ノーマル (normal)	1.vvm
九州そら (Kyushu Sora)	16 ノーマル (normal), 15 あまあま (sweet), 17 セクシー (sexy), 18 ツンツン (prickly), 19 ささやき (whisper)	2.vvm, 5.vvm
もち子さん (Mochikosan)	20 ノーマル (normal), 66 セクシー／あん子 (sexy/Anko), 77 泣き (crying), 78 怒り (angry), 79 喜び (joy), 80 のんびり (relaxed)	15.vvm
剣崎雌雄 (Kenzaki Mesuo)	21 ノーマル (normal)	4.vvm
WhiteCUL	23 ノーマル (normal), 24 たのしい (happy), 25 かなしい (sad), 26 びえーん (crying)	8.vvm
後鬼 (Goki)	27 人間ver. (human), 28 ぬいぐるみver. (plush), 87 人間（怒り）ver. (human, angry), 88 鬼ver. (demon)	7.vvm, 16.vvm
No.7	29 ノーマル (normal), 30 アナウンス (announcer), 31 読み聞かせ (storytelling)	6.vvm
ちび式じい (Chibishikiji)	42 ノーマル (normal)	10.vvm
櫻歌ミコ (Ouka Miko)	43 ノーマル (normal), 44 第二形態 (second form), 45 ロリ (childlike)	11.vvm
小夜/SAYO (Sayo)	46 ノーマル (normal)	15.vvm
ナースロボ＿タイプＴ (Nurserobo Type T)	47 ノーマル (normal), 48 楽々 (relaxed), 49 恐怖 (fear), 50 内緒話 (secret talk)	11.vvm
†聖騎士 紅桜† (Horinaito Benizakura)	51 ノーマル (normal)	12.vvm
雀松朱司 (Wakamatsu Akashi)	52 ノーマル (normal)	12.vvm
麒ヶ島宗麟 (Kigashima Sourin)	53 ノーマル (normal)	12.vvm
春歌ナナ (Haruka Nana)	54 ノーマル (normal)	13.vvm
猫使アル (Nekotsuka Aru)	55 ノーマル (normal), 56 おちつき (calm), 57 うきうき (cheerful), 110 つよつよ (strong), 111 へろへろ (exhausted)	13.vvm, 21.vvm
猫使ビィ (Nekotsuka Bi)	58 ノーマル (normal), 59 おちつき (calm), 60 人見知り (shy), 112 つよつよ (strong)	13.vvm, 21.vvm
中国うさぎ (Chugoku Usagi)	61 ノーマル (normal), 62 おどろき (surprised), 63 こわがり (scared), 64 へろへろ (exhausted)	3.vvm
栗田まろん (Kurita Maron)	67 ノーマル (normal)	14.vvm
あいえるたん (Aierutan)	68 ノーマル (normal)	14.vvm
満別花丸 (Manbetsu Hanamaru)	69 ノーマル (normal), 70 元気 (energetic), 71 ささやき (whisper), 72 ぶりっ子 (cutesy), 73 ボーイ (boy)	14.vvm
琴詠ニア (Kotoyomi Nia)	74 ノーマル (normal)	14.vvm
Voidoll	89 ノーマル (normal)	17.vvm
ぞん子 (Zonko)	90 ノーマル (normal), 91 低血圧 (groggy), 92 覚醒 (awakened), 93 実況風 (commentator)	18.vvm
中部つるぎ (Chubu Tsurugi)	94 ノーマル (normal), 95 怒り (angry), 96 ヒソヒソ (hushed), 97 おどおど (timid), 98 絶望と敗北 (despair and defeat)	18.vvm
離途 (Rito)	99 ノーマル (normal), 101 シリアス (serious)	19.vvm
黒沢冴白 (Kurosawa Kohaku)	100 ノーマル (normal)	19.vvm
ユーレイちゃん (Yureichan)	102 ノーマル (normal), 103 甘々 (sweet), 104 哀しみ (sorrow), 105 ささやき (whisper), 106 ツクモちゃん (Tsukumo-chan)	20.vvm
東北ずん子 (Tohoku Zunko)	107 ノーマル (normal)	21.vvm
東北きりたん (Tohoku Kiritan)	108 ノーマル (normal)	21.vvm
東北イタコ (Tohoku Itako)	109 ノーマル (normal)	21.vvm
あんこもん (Ankomon)	113 ノーマル (normal), 114 つよつよ (strong), 115 よわよわ (weak), 116 けだるげ (languid), 117 ささやき (whisper)	22.vvm, 23.vvm
夜語トバリ (Yogatari Tobari)	118 ノーマル (normal), 119 明るい (bright), 120 哀しみ (sorrow), 121 呆れ (exasperated)	24.vvm
暁記ミタマ (Akatsuki Mitama)	122 ノーマル (normal), 123 怒り (angry), 124 哀しみ (sorrow), 125 ささやき (whisper)	24.vvm
里石ユカ (Satoishi Yuka)	126 つぼみ (Tsubomi)	24.vvm

GPU

To use CUDA, download the CUDA build of ONNX Runtime and the CUDA libraries into the same directory:

./download -o ~/voicevox_core --devices cuda --only onnxruntime additional-libraries

The ONNX Runtime CUDA provider needs to find the CUDA and cuDNN libraries in additional_libraries. Call LoadAdditionalLibraries before LoadOnnxruntime, then set the acceleration mode when you create the synthesizer:

voicevox.LoadAdditionalLibraries("voicevox_core/additional_libraries")
ort, _ := voicevox.LoadOnnxruntime("voicevox_core/onnxruntime/lib")

opts := voicevox.DefaultInitializeOptions()
opts.AccelerationMode = voicevox.AccelerationModeGPU
syn, err := voicevox.NewSynthesizer(ort, ojt, opts)

NewSynthesizer returns an error if the GPU can't be used. Call syn.IsGPUMode() to see which mode is active, and ort.SupportedDevices() to see which devices the ONNX Runtime build supports.

go run ./examples/hello -data ~/voicevox_core -gpu

Tests

VOICEVOX_LIB=~/voicevox_core/c_api/lib VOICEVOX_DIR=~/voicevox_core go test ./...

VOICEVOX_LIB is required. Tests that need voice models are skipped when VOICEVOX_DIR is not set. To also run the GPU test, set VOICEVOX_TEST_GPU=1.

Credits

The VOICEVOX voice model terms require a credit for each character you use, for example "VOICEVOX:ずんだもん". Check the terms for each character.

License

MIT. VOICEVOX core, its voice models and its ONNX Runtime each have their own licenses and terms.