Kokoro

Kokoro TTS

An 82M-parameter open model from Hexgrad that delivers studio-quality speech at nearly 100x real-time.

Bhala for 5,000 characters limit

Ukufaka umbhalo wakho kumathegi we-SSML ukulawula okucacile:

<speak><prosody rate="slow">Slow speech</prosody></speak>

Amathegi amamodeli akhethiwe aqonda - chofoza ukuwasusa kusihloko sakho lapho kwenzeka khona:

Le modeli ifunda umbhalo ojwayelekile, ngakho amathegi e-inline akhohlwa. Ukwenza umbono osekelwe kumathegi, shintsha kwimodeli ebonisa umbono njenge-Orpheus noma i-Bark.

Chaza ukuchaza okujwayelekile (igama = ukuchaza):

-12 +12
0.5x 2.0x
Imahhala ne-Piper, VITS, MeloTTS
Umsindo wakho okhiqizwe uzovela lapha. Khetha imodeli, ngenisa umbhalo, bese uchofoza Ukukhiqiza.
Umsindo wakhiwa ngokuphumelelayo
0:00
Layisha phezulu umsindo Layisha phezulu.srt Isixhumanisi siphele ngehora le-24
Isikhashana esimahhala: ukusetshenziswa komuntu siqu. Ilayisense yebhizinisi kusuka ku-$5/mo
Yenza lokhu kube umsindo wakho Uhlu lwezinhlamvu
Uthanda i-TTS.ai? Ncoma abangane bakho!

Ngo Kokoro

Kokoro, built by Hexgrad, is a deliberately tiny 82-million-parameter model that punches far above its size class. It uses a StyleTTS + ISTFTNet architecture and was trained on roughly 1,200 hours of speech, yet generates audio close to 100x faster than real-time on a GPU while staying remarkably natural and expressive. It covers English, Japanese, Chinese, French, Italian, Portuguese, Spanish, and Hindi with a varied set of expressive voicepacks, and supports streaming. The combination of small footprint, low latency, and a permissive license has made Kokoro one of the most popular free models for high-volume and streaming use — it carries the largest share of traffic on TTS.ai.

Okungcono kakhulu: High-quality TTS with minimal latency, streaming applications

Khangela konke Kokoro izizwi

Ngombono ocacile

Umthuthukisi
Hexgrad
Ilayisense
Apache 2.0
I-Tiger
free
Isivinini
fast
Ukuklona umsindo
Akukho
Izilimi
English, Japanese, Chinese, French, Italian, Portuguese, Spanish, Hindi
Amaphawu aphezulu
500

Kokoro izizwi

Adam

English
Ikhululekile Male

Alex

Spanish
Ikhululekile Male

Alex

Portuguese
Ikhululekile Male

Alpha

Japanese
Ikhululekile Female

Alpha

Hindi
Ikhululekile Female

Bella

English
Ikhululekile Female

Dora

Portuguese
Ikhululekile Female

Dora

Spanish
Ikhululekile Female

Emma (British)

English
Ikhululekile Female

George (British)

English
Ikhululekile Male

Gongitsune

Japanese
Ikhululekile Female

Heart

English
Ikhululekile Female

Isabella (British)

English
Ikhululekile Female

Lewis (British)

English
Ikhululekile Male

Michael

English
Ikhululekile Male

Nicola

Italian
Ikhululekile Male

Nicole

English
Ikhululekile Female

Omega

Hindi
Ikhululekile Male

Sara

Italian
Ikhululekile Female

Sarah

English
Ikhululekile Female

Siwis

French
Ikhululekile Female

Sky

English
Ikhululekile Female

Xiaobei

Chinese
Ikhululekile Female

Xiaoni

Chinese
Ikhululekile Female

Xiaoxiao

Chinese
Ikhululekile Female

Yunjian

Chinese
Ikhululekile Male

Kokoro Imibuzo ebuzwa kaningi

Yes. Kokoro is released under the Apache 2.0 license and sits in the free tier, so it can be used commercially at no cost.

Kokoro is only 82M parameters and uses an efficient StyleTTS + ISTFTNet design, letting it generate audio nearly 100x faster than real-time on a GPU with just ~1.5GB VRAM while staying natural-sounding.

Kokoro supports English, Japanese, Chinese, French, Italian, Portuguese, Spanish, and Hindi, each with its own expressive voicepacks. It does not support voice cloning.
← Zonke izizwi