JVS Corpus

by Saruwatari Lab (U. Tokyo)

30h Japanese versatile multi-speaker corpus — TTS + speaker-modeling baseline.

TL;DR

30h Japanese versatile multi-speaker corpus — TTS + speaker-modeling baseline.

Best for japanese multi-speaker TTS, speaker adaptation, voice conversion. Pricing: research-only.

Category
Open source
License
—
Stars
—
Last push
—
Pricing
research-only
Platforms
Web

What it is

JVS (Japanese Versatile Speech) is a 30-hour multi-speaker corpus with parallel/non-parallel/whisper/falsetto styles across 100 speakers. Standard JP TTS + voice-modeling baseline. Non-commercial license.

Best for: Japanese multi-speaker TTS, speaker adaptation, voice conversion.
Watch out for: Custom non-commercial license · 100 speakers · parallel + non-parallel + whisper + falsetto styles. Cite: Takamichi et al., 2019.

Install / use

https://sites.google.com/site/shinnosuketakamichi/research-topics/jvs_corpus

Features

Speaker diarizationNo
Word-level timestampsNo
Streaming / real-timeNo
Languages supported1
HIPAA eligibleNo

JVS Corpus vs Whipscribe

FeatureJVS CorpusWhipscribe
CategoryOpen sourceTranscription APIs
Pricingresearch-onlyfree beta
Speaker diarizationNoYes
Word timestampsNoYes
StreamingNoNo
Languages199
PlatformsWebWeb, API, MCP

Alternatives to JVS Corpus

OpenAI Whisper
OpenAI

The reference open-source multilingual ASR model from OpenAI.

OSS · MIT ★ 98.1k
whisper.cpp
Georgi Gerganov

C/C++ port of Whisper — runs on anything, from a Raspberry Pi to Apple Silicon.

OSS · MIT ★ 48.8k
faster-whisper
SYSTRAN

4× faster than reference Whisper using CTranslate2 — production sweet spot.

OSS · MIT ★ 22.3k

Whipscribe is a managed faster-whisper + whisperX service. If you want transcripts without running infrastructure, paste a URL or drop a file in the form below — you'll have a transcript in seconds.