Isebenza nge Chatterbox, Kokoro & Whisper

I-Ultra
Ukukhuluma kulokho okukhuluma
I-AI Platform

Uhluze noma yiluphi ucingo emaminithini, uguqule umbhalo ube ulwimi, ubhale umsindo futhi uguqule ucingo olulodwa lube olulodwa, konke ngemodeli yomthombo ovulekile.

Akukho khadi le-credit elidingekayo 10 credits free Ukungena kwe-API kufakwe
Uhlu lwezinhlamvu
Igama elifanayo, umsindo ohlukile

Ukuzwa isiqephu esifanayo ngezizwi ezimbili ezihlukene

Uhlu lwezinhlamvu

Isebenza ngemodeli ye-open-source esezingeni eliphakeme

Qwen2.5 Chatterbox Kokoro 82M Whisper Large v3 OpenVoice v2

Konke okudingayo ngezwi AI

Amathuluzi aphelele kuwo wonke amathuluzi wokukhuluma

Indlela esebenza ngayo

Izinqumo ezintathu ezilula zokushintsha umsindo

1

Layisha umsindo

Layisha phezulu isifanekiso somsindo (imizuzu engu-10-30) noma urekhode ngqo kwi-browser yakho. Sixhasa i-MP3, i-WAV, i-FLAC, nezinye.

2

Inqubo nge AI

Amaseva ethu we-GPU asebenza umsindo wakho ngemodeli yomthombo ovulekile njenge-Chatterbox, Kokoro, Whisper ne-OpenVoice.

3

Layisha phezulu noma udlulise

Thola izimpendulo zakho ngokushesha. Layisha ngezansi ngefomethi eziningi noma udlulise ngokuqondile nge-API yethu yesikhathi sangempela.

Isebenza ngezindlela ezingcono kakhulu zomthombo ovulekile

Sisebenzisa futhi sisiza ucwaningo lwe-AI olusezingeni eliphakeme

OpenVoice v2

Injini yokushintsha umsindo ekhethiwe kanye nemodeli yokuklonya

  • Ukushintsha umbala wethoni
  • Igcina amagama nesikhathi
  • Ilayisense le-MIT
Chatterbox

Ukuklonya kwezwi, ukudubula kwezwi kanye nokuguqulwa kwezwi lethu elizenzakalelayo (Chatterbox VC)

  • Ukuklonywa kwe-zero-shot
  • 23 languages
  • Ukusetshenziswa kwebhizinisi kuvunyelwe
Kokoro 82M

I-text-to-speech esheshayo, ejwayelekile

  • 54 voices
  • 8 languages
  • Ilayisense le-Apache 2.0
Whisper Large v3

Ukudluliswa kwezilimi eziningi

  • Ukuthola ulwimi ngokuzenzakalela
  • 100 languages
  • Igama lesikhathi

Ifakwe ngezinye izimo zokusebenziseka

Ukusuka ekuthuthukiseni isihloko kuya ekusebenziseni ibhizinisi

Ukwenza okuqukethwe

Dala izizwi eziphezulu zevidiyo, amapodcast, amabhuku omsindo, kanye nesimo semidiya yomphakathi ngokulinganisela.

Usizo lwamakhasimende

Yakha izisebenzi zokukhuluma ezisho ukuthi zizwakala kahle futhi zingaphatha ukuxhumana okunzima.

Ukudweba

Ifaka izithameli ezihlukahlukene ezikhuluma izilimi eziningi ngenkathi ugcina umsindo womuntu okhulumayo.

Ufinyelela

Sinikeze umsindo labo abalahlekelwa umsindo wabo. Dala isingeniso sokukhuluma esizimele.

Imidlalo nemidlalo

Dala izizwi ze-NPC ezinamandla, izincoko zomuntu, kanye nezimo ezixhumene.

Ukufundisa

Dala okuqukethwe okuhlobene nokufunda ngemibhalo ezwakalayo.

Umthuthukisi we-API

Yakha nge-API yethu

I-REST API elula yethuluzi ngalinye: thumela isicelo nge-API yakho, bese ukhetha imiphumela.

  • I-RESTful API

    Izinkomba zesiphelo ze-HTTP ezilula zo zonke izici

  • Ama-Webhooks

    Imisebenzi yokudlulisa inga-POST imiphumela yayo ku-URL yakho uma iqediwe

Bona amadokhumende we-API
POST /v1/voice-clone/
curl -X POST https://api.speechtospeechai.com/v1/voice-clone/ \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "reference_audio=@sample.mp3" \
  -F "text=Hello, this is my cloned voice." \
  -F "model=chatterbox"

# Response: a job to poll
{"uuid": "3f2c...", "status": "queued"}

curl "https://api.speechtospeechai.com/v1/speech/results/?uuid=3f2c..." \
  -H "Authorization: Bearer YOUR_API_KEY"

{"status": "completed", "result_url": "https://.../output.wav"}

Intengo elula, ecacile

Qala ngokukhululekile, ulinganisele njengoba ukhula

Ikhululekile
$0/izinsuku
  • 10 credits/month
  • Zonke izithuluzi zifakazelwa
  • Ukufinyelela kwe-API
Qala
Isiqalisi
$9/izinsuku
  • 100 credits / Inyanga
  • Zonke izithuluzi zifakazelwa
  • Ukufinyelela kwe-API
Khetha isiqalisi
I-Pro
$29/izinsuku
  • 500 credits / Inyanga
  • Zonke izithuluzi zifakazelwa
  • Ukufinyelela kwe-API
Khetha i-Pro
I-Enterprise
$99/izinsuku
  • 2000 credits / Inyanga
  • Zonke izithuluzi zifakazelwa
  • Ukufinyelela kwe-API
Khetha i-Enterprise

Imibuzo ebuzwa kaningi

Konke odinga ukukwazi mayelana Speech to Speech AI

Ukuklonya kwezwi kusetshenziswa i-AI ukubheka isampula sezwi lamuntu (okungukuthi, amasekondi angama-10-30) futhi kwenziwa imodeli yedijithali engakwazi ukukhuluma noma yisiphi isibhalo kulolu zwi. Iphrojekthi yethu isebenzisa amamodeli asezingeni eliphakeme njenge-Chatterbox ne-OpenVoice ukufinyelela emiphumela ezwakalayo.

Sixhasa isiNgisi, isiSpanishi, isiFulentshi, isiJalimane, isiChinese, isiJapanese, isiKorea, isiArabhu, isiPutukezi, isiRussia, isiItalian, isiDutch, isiPolish kanye nezinye eziningi. Izimo zethu eziningi-zinsiNgisi zingaklona izizwi futhi zikhiqize ukukhuluma nganoma yisiphi sezilimi.

Yebo! Sinikeza isigaba esimahhala esinama-credits angama-10 ngenyanga, esifaka ukufinyelela okugcwele kuzo zonke izici: ukuklonya kwezwi, umbhalo-ku-ukukhuluma, ukukhuluma-ku-ukukhuluma, nokuningi. Akukho ikhadi le-credit elidingekayo ukubhalisa. Ukukhuphuka nganoma yisiphi isikhathi ukuze uthole ama-credits amaningi.

Sisebenzisa amamodeli avulekile: iChatterbox ne OpenVoice v2 ukuklonya umsindo, iKokoro, iPiper, iVITS ne Bark ukuhlela umbhalo ube kukhuluma, iWhisper Large v3 ukudlulisa, iChatterbox VC ukushintsha umsindo (iOpenVoice v2 ihlala ikhethwa), iChatterbox ukuklama umsindo ku-AI dubbing kanye neQwen2.5 ukuphendula ku-voice chat. Zonke zisebenza kumaseva ethu we-GPU.

Kulungile! Zonke izinhlelo zethu ezikhokhelwa zifaka phakathi ilungelo lokusetshenziswa kokuthengiswayo. Ungasebenzisa umsindo okhiqizwe ngevidiyo, amapodcast, ama-apps, imidlalo, usizo lwamakhasimende, nanoma yisiphi isisebenziso sokuthengiswayo. Qinisekisa ukuthi unelungelo lokuklonya noma yiziphi izizwi osebenzisayo.

Ucingo lwezingxoxo lukuvumela ukuthi uxoxe nosizo lwe-AI. I-browser yakho iguqula ukukhuluma kwakho kube sihloko, i-Qwen2.5 (imodeli yesilimi esivulekile esisebenza ku-GPUs zethu) ibhala impendulo kumuntu okhetha, futhi ucingo lwe-Piper luyifunda ngokuzwakalayo. Isebenza ngokushintshana, ngesiNgisi.

Ukulungele ukushintsha izisebenziso zakho zomsindo?

Xhumana namakhulu ochwepheshe abathuthukisi nabahleli besebenzisa i-Speech to Speech AI.