OpenVoice

Ultra

Ukulungiswa kwezwi okuzenzakalelayo nge-granular tone control

Moderate Isivinini
Very Good Ubunjani
Yebo Ukuklonya
10 Izilimi

Ngo OpenVoice

of the voice and apply it to the speaking style. OpenVoice is a powerful voice clone model that allows you to clone voices from a variety of sources, including voices from a variety of sources, including voices from a variety of sources, including voices from a variety of sources, including voices from a variety of sources, including voices from a variety of sources. You can then clone the

Izici ezibalulekile

Ukuklonya okuzenzakalelayo

Uhlu lwezinhlamvu

Ukulawula umsindo

Sebenzisa amathoni amnandi, abuhlungu, abuhlungu, athakazelisayo, noma aphuthumayo.

Ukudluliswa kwesimo

Yakha isibonakaliso somsindo kusuka kuhlobo lokukhuluma ukuze kube lula ukusisebenzisa.

isi-Latin

Sebenzisa izizwi eziklonyelelwe phakathi kwezilimi ezahlukene.

Ukucubungula okukhawulelwe

Ukucabanga okunembile kokukhiqizwa kwezwi okusheshayo.

Umsuka ovulekile

I-MIT ilayisense izicelo zokuhweba.

Sebenzisa izimo

Isihloko esithakazelisayo Umsebenzi wokuzijabulisa Imidlalo exhunywe Ukukhuluma incwadi yomsindo Amavidiyo e-Marketing Ama-Asistents abonakalayo

Indlela yokusetshenziswa OpenVoice

  1. 1

    Ubhalise mahhala noma hlola idemo

    Dala i-akhawunti emahhala ye-TextToSpeechAI ukuze uthole ama-credits wokuqamba, noma sebenzisa ikhasi lokubonisa ukulalela i-OpenVoice ngaphambi kokufaka. Akukho GPU noma ukufaka okudingayo - konke kusebenza kumaseva ethu.

  2. 2

    Khetha i-OpenVoice bese ufaka umsindo obhekiswe kuwo

    Khetha i-OpenVoice engine, bese ufaka imizuzwana embalwa yokuxhumana okuhlanzekile ukuze uklonyelise ngokushesha umsindo ofuna ukuwufinyelela. I-OpenVoice iqoqa ukuxhumana komsindo ukuze ukwazi ukuyisebenzisa kabusha nganoma iyiphi incwadi nethoni.

  3. 3

    Faka umbhalo wakho

    Bhala noma chofoza isikripthi ofuna ukusikhuluma emlanjeni ohlonywe. I-OpenVoice isekela amagama angaphezu kuka-10 kanye nokuthunyelwa kwe-cross-language, ngakho-ke ungabhala ulwimi oluhlukile kunesiqephu esibhekiswe kuso.

  4. 4

    Khetha uhlobo lwesithonjana bese udala

    Khetha enye yezindlela ezingu-9 ze-OpenVoice tone - iphutha, ethandekayo, ejabulisayo, ethokozisayo, ebuhlungu, ekhohlisayo, ekhathazekile, ekhalayo, noma ephuthumayo - bese udala. Umsindo ofanayo ohlobene uzokhuluma ngokunikeza okunengqondo.

  5. 5

    Layisha phezulu noma sebenzisa i-API

    Layisha ngezansi umsindo wakho njenge MP3, WAV, noma OGG, noma usebenzise ukukhishwa okuzenzakalelayo nge-TextToSpeechAI API ngokudlulisa umsindo wakho oklonyeliwe nesimo se-tone kunoma yisiphi isicelo.

OpenVoice I-API

Yenza ulwimi ngokuzenzakalela usebenzisa i-TextToSpeechAI REST API.

curl -X POST "https://api.texttospeechai.com/v1/generate/" \
  -H "Authorization: Bearer YOUR_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "I\u002DOpenVoice ingakhuluma nganoma iyiphi into \u002D ejabulisayo, ebuhlungu, noma ngisho nokuphuza.",
    "voice": "en_US-lessac-medium"
  }'

Imibuzo ebuzwa kaningi

OpenVoice iyimodeli ethuthukisiwe yokubhala-kuya-kwezwi nokukhuluma ehlukanise ngokuhlukile ukuxhumana kwezwi nohlobo lokukhuluma. Le ndlela ikuvumela ukuthi uhlukanise ukukhuluma bese ufaka izithonjana ezahlukene zengqondo ngaphandle kokudinga ukuxhumana okusha kwezwi ngengqondo ngayinye. Ifakwe ukubonisa, ukulawulwa kokukhiqizwa kwezwi.

Yebo, i-OpenVoice isebenza ngokuzenzakalela ukuklonya umsindo kusuka emizuzwini emincane kuphela ye-audio ebhekiswe kuyo - akukho kusebenza kokuzivocavoca okudingayo. Uma umsindo uthathwa, i-OpenVoice ingasebenzisa kabusha lo msindo nganoma iyiphi incwadi noma iyiphi indlela ye-tone okhetha yona.

OpenVoice uses a two-stage architecture that splits base speech synthesis from tone conversion. After cloning a voice, you can apply any of 9 tone styles - default, friendly, cheerful, excited, sad, angry, terrified, shouting, or whispering - and the same cloned voice speaks differently based on your chosen tone without re-recording.

OpenVoice ixhasa izimo ezingu-9 zokukhuluma: iphutha, elihle, elijabulisayo, elithakazelisayo, elibuhlungu, elikhohlisayo, elikhathazekile, elishayayo, neliphuthumayo. Isimo ngasinye siguqula ukuthunyelwa kwemizwa ngenkathi sigcina isikhulumi esihlonyelwa, sikunika ukulawula okuncane-e-granular ngalokho okufundeka icala.

I-OpenVoice ikhona ngaphansi kwelayisense ye-MIT, ngakho-ke imahhala ukusetshenziswa kwezokuhweba. Njengokwesinye isifanekiso sokuklonya, qiniseka ukuthi unelungelo elifanele nganoma yisiphi isiqophi esiklonywe ngezinhloso zokuhweba.

I-OpenVoice isekela amagama angaphezu kwama-10 kufaka phakathi isiNgisi, isiChinese, isiJapane, isiKorea, kanye neminye amagama ase-Europe. Inikeza futhi ukuklonywa kwe-cross-language, ngakho-ke ungaklonya umsindo ku-language eyodwa futhi uyikhulume ngokuvamile ku-other.

I-OpenVoice inejubane lokukhishwa eliphakathi, livame ukuveza imvume emaminithini angama-2-4 ku-GPU. Umphumela wokusebenza ulungile kakhulu, nokwehla kwezwi okucacile nokudluliswa kwethoni okugcina isikhulumi sikhona ngenkathi sishintsha ngokuqinisa ukuthunyelwa kwemizwa.

OpenVoice idinga 6-8GB ye VRAM ngokuya ngenani leqembu nethonya lokushintshana. Isebenza kahle kuma-GPUs aphakathi nendawo kuya ephezulu, futhi ku-TextToSpeechAI konke lokhu kuphathwa kumaseva ethu ngakho-ke awudingi noma yiziphi izinsimbi zangaphakathi.

OpenVoice yinjini ye-Ultra-tier, ethengiswa nge-50 credits ngamagama angu-1000. I-Ultra-tier ibonisa ukulawulwa kwethoni esezingeni eliphakeme kanye nokusebenza okungeziwe okudingayo ukuklonya kanye nesimo-sokuguqulwa kwe-pipeline.

OpenVoice ihlukile ngenxa yethoni yayo kanye nesimo sokulawula: ungathatha umsindo owodwa ohlobene futhi uphinde unikezele njengomnandi, obuhlungu, obuhlungu, noma oxhumanayo. F5-TTS ishesha futhi iyinjini yethu ejwayelekile yokuhlobana yokuziphatha, ukukhuluma okungenalutho. Khetha iOpenVoice uma ufuna ukulawula isimo sokuziphatha, futhi F5-TTS uma ufuna ukuklonya okusheshayo okujwayelekile.

Dala umsindo ohlobene ngokufaka umsindo obhekiswe kuwo, bese uchaza uhlobo lwe-tone ku-API yakho. I-API ifaka umsindo othandekayo okhethiwe ku-voice ohlobene ngokuzenzakalela futhi ibuyisela umsindo ku-MP3, WAV, noma i-OGG format.

Yebo. Bhala i-akhawunti emahhala ye-TextToSpeechAI ukuze uthole ama-credits wokuqagela futhi uzame ukuklonya kwe-OpenVoice kanye nokulawula umsindo, noma sebenzisa ikhasi lokubonisa kuqala. Akukho silungiselelo sasekhaya - khuphula isiqephu esibhekiswe, khetha umsindo, futhi udale kwi-browser.

Technical Specs

  • Generation Speed Moderate
  • Output Quality Very Good
  • Voice Cloning Supported
  • Languages 10
  • GPU VRAM 3-6GB
  • Credits/1000 chars 50

Try OpenVoice Now

Generate your first audio free. No credit card required.

Start Free