Okubhaliweyo ukuya kuSpeechName Abasebenzi Umculo Uthetha ukuba UmbhaloName I-Voice Cloning Uyilo lwesandi I-Marketplace Okuninzi

I-Free AI Okubhaliweyo ukuya kuSpeechName

20+ iimodeli zomthombo ovulekileyo, 107+ ii-voices, 32+ iilwimi. Akukho akhawunti ifunekayo.

1K+
abavelisi
2K+
iindidi
20+
Imodeli ye-AI
107+
iilizwi
0/500 Iimpawu Ekhululekileyo
5,000 chars per generation 15,000 iimpawu ezikhululekileyo Akukho khadi letyala Ukusetyenziswa kwentengiso
Sign up free
0:00 / 0:00
Layisha ezantsi Ikhonkco liphelelwe lixesha kwiyure ezi-24
Like TTS.ai? Tell your friends!
Advanced TTS tool with 20+ models, speed control & file export

Yonke into oyifunayo kwi Voice AI

Izixhobo ezingaphezu kwe-30 ezixhaswa ziimodyuli ze-AI ezivulekileyo

Okubhaliweyo ukuya kuSpeechName

20+ iimodeli ze-AI, 100+ ilizwi, 30+ ulwimi

Uthetha ukuba UmbhaloName

I-Whisper, Faster Whisper, SenseVoice. 99 languages

Uguqulelo lwesandi

Uguqulelo lwesandi & lwevidiyo kumbhalo nge-AI

I-Voice Cloning

Ukwenza i-clone yesandi ngaphaya kwemizuzu emi-5 yesandi

Incoko YelizwiName

Ukuthetha ne-AI ngelizwi eliqhelekileyo ngexesha elibonakalayo

I-AI Agents

I-AI enamandla elizwi malunga nenkonzo yomthengi & ngaphezu

I-AI Music

Yenza umculo wokuqala ukusuka kwiinkcazelo zombhalo

Umtshintshi WesandiName

Iguqula uphawu lwesandi, udidi, isivakalisi

Umxhasi WesandiName

Susa ingxolo, yongeza ucacileyo, upscale umgangatho

Umcimi Wengoma

Susa iingoma okanye iinkwenkwezi ezisuka kwingoma

Umqalisi Wesandi

Yahlula iingoma zibe zii-drums, i-bass, iingoma, nezinye

Uguqulelo lomsebenzi kolunye ulwimi

Uguqulelo lomsebenzi kolunye ulwimi

Ukuthetha-thethana

Tshintsha ilizwi, iimvakalelo, ulwimi, uhlobo

Um_susa we-Echo

Susa i-echo ne-reverberation yegumbi

Isitshixo & BPM Umfumani

Fumana isitshixo somculo kunye ne-tempo ngokuzenzekelayo

Umguquli WesandiName

Guqula MP3, WAV, FLAC, OGG, M4A

Umfundi Woxwebhu

Linganisa amanqaku, ii-PDFs, kunye namaxwebhu

I-Audio StudioName

Iincwadi ezinesandi kunye neepodcast ezinesahluko esininzi

Umenzi Wepodcast

Iincoko ze-AI ezibini ezinomququzeleli ukusuka nakweyiphi na into

Iziphumo zesandi

Yenza iziphumo zesandi ukusuka kumyalezo okhawulezayo

I-Batch TTS

Guqula amawaka ombhalo nge-CSV upload

I-Dubbing Studio

Dlala ividiyo enesandi kwiilwimi ezingama-30+

& Isixhobo Somcimi

Dibanisa i-TTS nakweyiphi na iwebhusayithi ngomgca omnye

20+ Iimodeli zesandi ze-AI

Uluhlu olupheleleyo lweemodeli ze-TTS ezivulekileyo kwinkqubo enye

KokoroKokoro

Free

Imodeli elula ye-82M yeparameter enika ukuthetha kwestudio-umgangatho okhawulezayo okhawulezayo.

Iiparamitha ze-82M Ekhawulezayo kakhulu Ilizwi elithethayo Ulwimi oluninzi Inkxaso ye-Streaming
Fast · 1.5GB VRAM Zama kwakhona

PiperPiper

Free

Inkqubo yendlela yokusebenza yokubhala- ukuya- ku- kuthetha ekhawulezayo, engaphakathi kwindlela yokusebenza yokubhala- ukuya- ku- kuthetha eyenziwe kakuhle ye Raspberry Pi nezixhobo ezifakwe ngaphakathi. Name

CPU- elungele Ukuphuma ngaphandle kwenethiwekhi kunokwenzeka 100+ izithethi 30+ Iilwimi Inkxaso ye SSML
Fast · 0 (CPU only) VRAM Zama kwakhona

VITSVITS

Free

I-encoder eguqukayo elawulwa ngokuzenzekelayo enemfundo echaphazelayo yesiphelo-se-siphelo sombhalo-ukuthetha.

Isiphelo-siphelo soxinzelelo I-Prosody eNtsha Uvavanyo olukhawulezayo Abathethi abaninzi
Fast · 1GB VRAM Zama kwakhona

MeloTTSMeloTTS

Free

Umbhalo ophezulu ephepheni ovela kwiilwimi ezininzi usuka kumazwi osebenza kwi CPU ngexesha elifutshane lokulibazisa.

I-CPU-ilungelelaniswe kakuhle Iilwimi ezininzi IsiNgesi-C Uluhlu lweeNkqubo Ixesha lokuphuma eliphantsi
Fast · 0.5GB (GPU optional) VRAM Zama kwakhona

BarkBark

Standard

Imodeli yombhalo-ukuya-kwisandi esekelwe kwi-transformer evelisa ukuthetha okunyanisekileyo, umculo, kunye neziphumo zesandi.

Iziphumo zesandi Uxolo/uxolo olukhulu Uhlobo lwengoma Abathethi abangaphezu kwe-100 Ulwimi oluninzi
Slow · 5GB VRAM Zama kwakhona

Bark SmallBark Small

Standard

Uguqulelo olusezantsi lwe Bark olunolwazi olukhawulezayo nokusetyenziswa okuphantsi kovimba wolwazi.

Iinkcukacha Ikhawuleza kune-Bark epheleleyo Ukuthetha ngokuzithandela Iilwimi ezininzi
Medium · 2GB VRAM Zama kwakhona

CosyVoice 2CosyVoice 2

Standard

I-Alibaba's scalable streaming TTS ene-human-parity naturalness kunye ne-zero-near latency.

Unikezelo Uklonelo lwe-zero-shot Iilwimi eziliqela Ulawulo lweemvakalelo I-Human-parity
Medium · 4GB VRAM Zama kwakhona

Dia TTSDia TTS

Standard

Imodeli yokuveliswa kwencoko yababini yesandi esininzi eyenza unxibelelwano oluqhelekileyo phakathi kwamasandi.

Isithethi esininzi Ukwenziwa kwencoko yababini Ukuthatha umjikelo oqhelekileyo Ukubonakaliswa kweemvakalelo 1.6B parameters
Medium · 4GB VRAM Zama kwakhona

Parler TTSParler TTS

Standard

Ichaza ilizwi ofuna ngayo kwilwimi oluqhelekileyo kwaye i-Parler ivelise ukuthetha okuhambelanayo.

Inkcazelo yeSandi Ulawulo lweelwimi zobuqu Ukwenza ilizwi elilula Akukho lizwi elimiselweyo elifunekayo
Medium · 4GB VRAM Zama kwakhona

GLM-TTSGLM-TTS

Standard

Ifumana umyinge womlinganiselo womonakalo wophawu olusezantsi phakathi kweemodeli ze-TTS ezivulekileyo.

Impazamo ephantsi Ukuphinda usebenzise ilizwi Uthelekiso lokuhamba I-Prosody eNtsha
Medium · 4GB VRAM Zama kwakhona

IndexTTS-2IndexTTS-2

Standard

I-TTS engapheliyo ene-fine-grained emotional control kunye nokubonisa okuphezulu.

Ulawulo lweemvakalelo I-Zero-shot Ii-Vectors zeMvakalelo Ukuthetha okuchazayo Ulawulo olunogranule encinci
Medium · 4GB VRAM Zama kwakhona

Spark TTSSpark TTS

Standard

Uklone lwelizwi le TTS ngeemvakalelo ezilawulwayo kunye nesitayile sokuthetha ngeempendulo.

Ukuphinda usebenzise ilizwi Ulawulo lweemvakalelo Ulawulo lwesimbo I-Prompt-based 5- imizuzwana yokuklona
Medium · 4GB VRAM Zama kwakhona

GPT-SoVITSGPT-SoVITS

Standard

Ilizwi elincinci-eliqhutywa lokuklonya i-TTS ephindayo nayiphi na ilizwi ukusuka kwimizuzu emihlanu kuphela yesandi.

5 imizuzwana yokukrola Ilizwi elidlalayo Ukufunda ngemizuzwana embalwa Ukuthembeka okuphezulu Iilwimi eziliqela
Slow · 6GB VRAM Zama kwakhona

OrpheusOrpheus

Standard

Imodeli ye-TTS evakalelwa ngamandla enqanaba lomuntu eqeqeshwe kwi-100K yeeyure zedatha yokuthetha.

Umgangatho wengqondo yomuntu 100K iiyure zoqeqesho Ukubeka ingqalelo ngokwendalo Ukuthetha okuchazayo
Medium · 4GB VRAM Zama kwakhona

ChatterboxChatterbox

Premium

Uhlobo olutsha lwesandi esingena-nto esifana nesandi esilawulwa ngumnqweno ovela kwiResemble AI.

Uklonelo lwe-zero-shot Ulawulo lweemvakalelo Ukuthembeka okuphezulu Unikezelo lwesimbo Uklonelo lwesampuli enye
Medium · 4GB VRAM Zama kwakhona

Tortoise TTSTortoise TTS

Premium

Umbhalo-ukuthetha-ngezwi oluninzi olujolise kwixabiso kunye noyilo oluya ezantsi ngokuzenzekelayo.

Ubunjani obuphezulu kakhulu Ilizwi elininzi Uyilo lweDALL-E Ukuphinda usebenzise ilizwi Ukuphinda-phinda okuzenzekelayo
Slow · 8GB VRAM Zama kwakhona

StyleTTS 2StyleTTS 2

Premium

Umgangatho womntu-umbhalo-ukuthetha-ukuthetha ngokusasaza isimbo kunye noqeqesho oluchaseneyo.

Umphakamo woMntu Uhlobo lokusasaza Uqeqesho oluchaphazela Utshintsho oluqhelekileyo Ukuthembeka okuphezulu
Medium · 4GB VRAM Zama kwakhona

OpenVoiceOpenVoice

Premium

Uklonelo lwesandi olukhawulezayo nolawulo oluthe kratya kwindlela, imvakalelo, nesiqhelo.

Uklonelo olukhawulezayo Uguqulelo lwesandi Ulawulo lweemvakalelo Ulawulo lwe-Accent Ulwimi oluninzi
Medium · 4GB VRAM Zama kwakhona

Qwen3 TTSQwen3 TTS

Standard

I-Alibaba's multilingual TTS enesandi sokukrola, ilizwi elimiselweyo, kunye noyilo lwesandi ukusuka kumbhalo.

Ukuphinda usebenzise ilizwi 9 ilizwi elichaziweyo phambi koshicilelo Uyilo lwesandi ukusuka kumbhalo Ulawulo lweemvakalelo Iilwimi
Medium · 7GB VRAM Zama kwakhona

Sesame CSMSesame CSM

Premium

Imodeli yokuthetha-thethana eyenza unxibelelwano oluqhelekileyo ngexesha elifanelekileyo kunye nengqondo.

Incoko Ixesha eliqhelekileyo Ukujika Isiqhagamshelanisi esezantsi Iiparamitha ze-1B
Slow · 8GB VRAM Zama kwakhona

KokoroKokoro Free

I-Kokoro yimodeli yombhalo-ukuthetha eneparameter ezili-82 ezili-million eyenza ungqubano oluhle ngaphezulu kweqela layo lobunzima. Nangona ubungakanani bayo buncinci, ivelisa ukuthetha okucacileyo nobucacileyo. I-Kokoro ixhasa ulwimi oluninzi oluquka isiNgesi, isiJaphani, isiTshayina, nesiKorea ngeendlela ezahlukeneyo zesandi ezicacileyo. Isebenza ngokukhawuleza kakhulu — ivelisa isandi esimalunga ne-100x ngokukhawuleza kunexesha elibonakalayo kwi-GPU.

Elungileyo ku: I-TTS esezingeni eliphezulu enexesha lokulibaziseka elincinci, iinkqubo zokudlulisa

Zama simahla

PiperPiper Free

I-Piper yinjini elula yombhalo-ukuthetha ephuhliswe yi Rhasspy esebenzisa i VITS kunye ne-larynx architectures. Isebenza ngokupheleleyo kwi CPU, iyenza ibe yindawo efanelekileyo yezixhobo zesiphelo, ulawulo lwasekhaya, kunye neenkqubo ezifuna i-offline TTS. Ngeelizwi ezingaphezu kwe-100 ezisuka kwiilwimi ezingaphezu kwe-30, i-Piper inikezela ngokuthetha okuziva ngathi kuqhelekanga kwisantya sexesha elibonakalayo nakwi-Raspberry Pi 4.

Elungileyo ku: Imboniselo yabucala ekhawulezayo, ufikelelo, kunye neenkqubo ezifakelweyo

Zama simahla

VITSVITS Free

VITS (I-Variation Inference ne-adversarial learning for end-to-end Text-to-Speech) yindlela efana ne-end-to-end TTS evelisa isandi esininzi esiqhelekileyo kunezikhokelo zenqanaba elinye. Isebenzisa i-variation inference ephuculweyo ngokuhamba okuqhelekileyo kunye nenkqubo yoqeqesho oluchaphazelayo, efumana ukuphuculwa okubalulekileyo kwindalo.

Elungileyo ku: Umbhalo-usuka-ku-ukuthetha osetyenziswa ngokubanzi nge-prosody eqhelekileyo

Zama simahla

MeloTTSMeloTTS Free

MeloTTS yi MyShell. ai yi TTS yelayibrari exhasa isiNgesi (iMelika, iBrithani, i-Indian, i-Australian), isiSpanyol, isiFrentshi, isiTshayina, isiJaphani, nesiKorea. Ikhawuleza kakhulu, iqhubekekisa umbhalo kwisantya esifutshane sexesha elibonakalayo kwi CPU kuphela. MeloTTS icwangciswe ukusetyenziswa kokwenza imveliso kwaye ixhasa zombini i CPU ne GPU inference.

Elungileyo ku: Iinkqubo zokuvelisa ezifuna i-TTS ekhawulezayo, eneelwimi ezininzi

Zama simahla

BarkBark Standard

Imodeli yombhalo-ukuya-kwisandi esekelwe kwi-transformer evelisa ukuthetha okunyanisekileyo, umculo, kunye neziphumo zesandi.

Umbhekisi phambili: Suno · Ilayisenisi: MIT

Zama kwakhona

Bark SmallBark Small Standard

Uguqulelo olusezantsi lwe Bark olunolwazi olukhawulezayo nokusetyenziswa okuphantsi kovimba wolwazi.

Umbhekisi phambili: Suno · Ilayisenisi: MIT

Zama kwakhona

CosyVoice 2CosyVoice 2 Standard

I-Alibaba's scalable streaming TTS ene-human-parity naturalness kunye ne-zero-near latency.

Umbhekisi phambili: Alibaba (Tongyi Lab) · Ilayisenisi: Apache 2.0

Zama kwakhona

Dia TTSDia TTS Standard

Imodeli yokuveliswa kwencoko yababini yesandi esininzi eyenza unxibelelwano oluqhelekileyo phakathi kwamasandi.

Umbhekisi phambili: Nari Labs · Ilayisenisi: Apache 2.0

Zama kwakhona

Parler TTSParler TTS Standard

Ichaza ilizwi ofuna ngayo kwilwimi oluqhelekileyo kwaye i-Parler ivelise ukuthetha okuhambelanayo.

Umbhekisi phambili: Hugging Face · Ilayisenisi: Apache 2.0

Zama kwakhona

GLM-TTSGLM-TTS Standard

Ifumana umyinge womlinganiselo womonakalo wophawu olusezantsi phakathi kweemodeli ze-TTS ezivulekileyo.

Umbhekisi phambili: Zhipu AI · Ilayisenisi: GLM-4 License

Zama kwakhona

IndexTTS-2IndexTTS-2 Standard

I-TTS engapheliyo ene-fine-grained emotional control kunye nokubonisa okuphezulu.

Umbhekisi phambili: Index Team · Ilayisenisi: Bilibili Model License

Zama kwakhona

Spark TTSSpark TTS Standard

Uklone lwelizwi le TTS ngeemvakalelo ezilawulwayo kunye nesitayile sokuthetha ngeempendulo.

Umbhekisi phambili: SparkAudio · Ilayisenisi: CC BY-NC-SA 4.0

Zama kwakhona

GPT-SoVITSGPT-SoVITS Standard

Ilizwi elincinci-eliqhutywa lokuklonya i-TTS ephindayo nayiphi na ilizwi ukusuka kwimizuzu emihlanu kuphela yesandi.

Umbhekisi phambili: RVC-Boss · Ilayisenisi: MIT

Zama kwakhona

OrpheusOrpheus Standard

Imodeli ye-TTS evakalelwa ngamandla enqanaba lomuntu eqeqeshwe kwi-100K yeeyure zedatha yokuthetha.

Umbhekisi phambili: Canopy Labs · Ilayisenisi: Llama 3.2 Community

Zama kwakhona

Qwen3 TTSQwen3 TTS Standard

I-Alibaba's multilingual TTS enesandi sokukrola, ilizwi elimiselweyo, kunye noyilo lwesandi ukusuka kumbhalo.

Umbhekisi phambili: Alibaba (Qwen) · Ilayisenisi: Apache 2.0

Zama kwakhona

ChatterboxChatterbox Premium

Uhlobo olutsha lwesandi esingena-nto esifana nesandi esilawulwa ngumnqweno ovela kwiResemble AI.

Ubunjani:

Zama kwakhona

Tortoise TTSTortoise TTS Premium

Umbhalo-ukuthetha-ngezwi oluninzi olujolise kwixabiso kunye noyilo oluya ezantsi ngokuzenzekelayo.

Ubunjani:

Zama kwakhona

StyleTTS 2StyleTTS 2 Premium

Umgangatho womntu-umbhalo-ukuthetha-ukuthetha ngokusasaza isimbo kunye noqeqesho oluchaseneyo.

Ubunjani:

Zama kwakhona

OpenVoiceOpenVoice Premium

Uklonelo lwesandi olukhawulezayo nolawulo oluthe kratya kwindlela, imvakalelo, nesiqhelo.

Ubunjani:

Zama kwakhona

Sesame CSMSesame CSM Premium

Imodeli yokuthetha-thethana eyenza unxibelelwano oluqhelekileyo ngexesha elifanelekileyo kunye nengqondo.

Ubunjani:

Zama kwakhona

CosyVoice 2CosyVoice 2

I-Alibaba's scalable streaming TTS ene-human-parity naturalness kunye ne-zero-near latency.

Iilwimi: en, zh, ja, ko, fr, de, it, es

Ilizwi lika-Clone

GLM-TTSGLM-TTS

Ifumana umyinge womlinganiselo womonakalo wophawu olusezantsi phakathi kweemodeli ze-TTS ezivulekileyo.

Iilwimi: en, zh

Ilizwi lika-Clone

IndexTTS-2IndexTTS-2

I-TTS engapheliyo ene-fine-grained emotional control kunye nokubonisa okuphezulu.

Iilwimi: en, zh

Ilizwi lika-Clone

Spark TTSSpark TTS

Uklone lwelizwi le TTS ngeemvakalelo ezilawulwayo kunye nesitayile sokuthetha ngeempendulo.

Iilwimi: en, zh

Ilizwi lika-Clone

GPT-SoVITSGPT-SoVITS

Ilizwi elincinci-eliqhutywa lokuklonya i-TTS ephindayo nayiphi na ilizwi ukusuka kwimizuzu emihlanu kuphela yesandi.

Iilwimi: en, zh, ja, ko

Ilizwi lika-Clone

ChatterboxChatterbox

Uhlobo olutsha lwesandi esingena-nto esifana nesandi esilawulwa ngumnqweno ovela kwiResemble AI.

Iilwimi: en

Ilizwi lika-Clone

Tortoise TTSTortoise TTS

Umbhalo-ukuthetha-ngezwi oluninzi olujolise kwixabiso kunye noyilo oluya ezantsi ngokuzenzekelayo.

Iilwimi: en

Ilizwi lika-Clone

OpenVoiceOpenVoice

Uklonelo lwesandi olukhawulezayo nolawulo oluthe kratya kwindlela, imvakalelo, nesiqhelo.

Iilwimi: en, zh, ja, ko, fr, de, es, it

Ilizwi lika-Clone

Qwen3 TTSQwen3 TTS

I-Alibaba's multilingual TTS enesandi sokukrola, ilizwi elimiselweyo, kunye noyilo lwesandi ukusuka kumbhalo.

Iilwimi: en, zh, ja, ko, de, fr, ru, pt, es, it

Ilizwi lika-Clone

Umbhekisi phambili-Okuqalayo

I-REST API ehambelana ne-OpenAI. Incopho enye yesiphelo, iimodeli ezingaphezu kwe-22. Inkxaso yosasazo lwezicelo zexesha elibonakalayo.

  • Ifomati ehambelana ne-OpenAI
  • Unikezelo lwe-TTS lweenkqubo zexesha elibonakalayo
  • Uqhubekeko lweqela lomsebenzi omkhulu
  • Isaziso se Webhook
Bonisa Uxwebhu lwe API
pip install ttsai npm install @ttsainpm/ttsai
Python 
from tts_ai import TTSClient

client = TTSClient(api_key="sk-tts-xxx")
audio = client.generate(
    text="Hello from TTS.ai!",
    model="kokoro",
    voice="af_bella",
)
client.save(audio, "output.mp3")

Ixabiso elilula, elicacileyo

Qala ngokukhululekileyo. Ubungakanani njengoko ukhula.

Ekhululekileyo

$0

15,000 characters

  • Kokoro, Piper, VITS, MeloTTS
  • Umda we-500 char
  • 3 gen/iyure (akukho akhawunti)
Ubhaliso simahla

Isiqalisi

$9/inyanga( ii)

500,000 characters/month

  • Zonke iimodeli ezingaphezu kwe-22
  • 100,000 chars per generation
  • I-Voice Cloning
Qala
Ethandwa Kakhulu

I-Pro

$29/inyanga( ii)

2,000 iikhredithi/inyanga

  • Yonke into kwisiqalisi
  • Ufikelelo lwe-API
  • Ukuqhubekeka okuphambili
Fumana i-Pro

Imisebenzi

$99/inyanga( ii)

10,000 iikhredithi/inyanga

  • Yonke into kwi-Pro
  • I-Bulk API
  • Ufolo oluphambili
Fumana iNkqubo

Bonisa zonke iinkqubo eziquka iipakeji zophawu →

Imibuzo ebuzwa rhoqo

TTS.ai yinkqubo yesandi ye-AI epheleleyo, enikezela ngeemodeli ezingaphezu kwe-22 zokubhala-ukuthetha, ukuclona kwelizwi, ukuthetha-ukubhaliweyo, kunye neezixhobo zesandi. Zonke iimodeli zivela kwi-open source ngaphandle kokuvula umboneleli.

Ewe! TTS.ai ibonelela ngemibhalo-ukuze-ithetha ngokukhululekileyo ngeemodeli zeKokoro, Piper, VITS, kunye neMeloTTS. Akukho akhawunti ifunekayo. Bhalisa ukuze ufumane amagama angama-15,000 asimahla kwaye ufike kuzo zonke iimodeli. Iinkqubo ezihlawulwayo ziqala kwi- $9/inyanga.

Ukusebenza ngokukhawuleza, sebenzisa iKokoro okanye iPiper. Ukusebenza kakuhle, zama iCosyVoice 2 okanye iStyleTTS 2. Ukwenza ilizwi lifana, sebenzisa iChatterbox okanye iGPT-SoVITS. Unxibelelwano, sebenzisa iDia TTS. Zama iimodeli ezininzi kumbhalo ofanayo ukuthelekiswa.

Ewe. I-REST API ehambelana ne-OpenAI ye-TTS, i-STT, ukuclona kwelizwi, kunye neezixhobo zesandi. Ifumaneka kwiPro ($29/mo) kunye ne-Enterprise ($99/mo) iinkqubo. Bona uxwebhu kwi-tts.ai/api/.

Ubunjani besandi buhluka ngokwemodeli. Iimodeli eziphezulu ezifana ne CosyVoice 2, StyleTTS 2, ne Chatterbox zivelisa ulwimi olunomgangatho ofanayo nolunobuntu obuqhelekileyo kunye novakalelo. Iimodeli ezikhululekileyo ezifana ne Kokoro zibonelela ngomgangatho olungileyo kwiziganeko ezininzi zokusetyenziswa.

I-TTS.ai ixhasa iilwimi ezingaphezu kwe-30 kwilayibrari yayo yemodeli. IsiNgesi sinokuxhasa imodeli ebanzi kakhulu, kodwa iimodeli ezifana ne-CosyVoice 2 ziquka isiTshayina, isiJapan, nesiKorea; i-GPT-SoVITS iphatha isiTshayina, isiJapan, isiKorea, nesiNgesi; kwaye i-MeloTTS ixhasa isiNgesi, isiSpanish, isiFrentshi, isiTshayina, isiJapan, nesiKorea.

Ewe. Zonke inkqubo ziqhutywa kwiseva yethu ekhethekileyo ye-GPU. Asigcinanga umbhalo wakho ongeniswe okanye isandi esiveliswe emva kokuthunyelwa. Iisampuli zesandi ezilayishwe phezulu zokuklonya zisetyenziswa kuphela kwintlanganiso yangoku kwaye azigcinwanga. Asiyi kudibana nedata yakho nabani na olandelayo okanye siyisebenzise ukuqeqesha iimodyuli.

Ewe. Zonke iiseshoni zesandi eziveliswe kwi-TTS.ai ziye zasetyenziswa ngokurhweba, kubandakanya i-YouTube videos, iipodcasts, iincwadi zesandi, ii-apps, izikhumbuzo, kunye neemveliso. Iimodeli zethu zivela kumbhalo ovulekileyo phantsi kwelayisensi ezivumelayo (MIT, Apache 2.0). Akukho lungelo lokushicilela okanye ukunikezelwa okufunekayo.

I-TTS.ai ivelisa isandi kwifomati ye-WAV ngokumiselweyo umgangatho ophezulu. Ungaguqula ukuya kwi-MP3, FLAC, OGG, okanye M4A usebenzisa isixhobo sethu esikhululekileyo sokuguqula isandi. I-API ixhasa ukukhankanya ifomati yakho ekhethiweyo yemveliso ngqo kwisicelo.

Layisha phezulu isampuli yesandi esezantsi (incinci njengemizuzwana emi-5) yelizwi ofuna ukulikhupha, emva koko ubhale nawuphi na umbhalo ukuvelisa ukuthetha kulo lizwi. Iimodeli ezinjenge Chatterbox, GPT-SoVITS, kunye ne CosyVoice 2 zixhasa ukulikhupha ulwimi. Ilizwi elikhuphiweyo lithatha into, isivakalisi, kunye nesitayile sokuthetha.

Iimodeli ezikhululekileyo (iKokoro, iPiper, iVITS, iMeloTTS) azidingi i-akhawunti kwaye zibiza uphawu olupheleleyo. Iimodeli eziqhelekileyo (2,000 uphawu/1K ingxelo) ziquka iBark, iCosyVoice 2, iF5-TTS, neDia. Iimodeli eziphezulu (4,000 uphawu/1K ingxelo) ziquka iOpenVoice, iChatterbox, iStyleTTS 2, neTortoise. Iimodeli ezihlawulwayo ngokubanzi zibonelela ngomgangatho ophezulu, iingoma ezininzi, kunye nemisebenzi engaphezulu njenge-cloning yelizwi.

Ewe. I-API ixhasa uqhubekeko lweqela lokuguqula ivolumu enkulu yombhalo kwilizwi. Thumela izicelo ezininzi kwaye ubuyisele iziphumo ngokuzenzekelayo usebenzisa umsebenzi we-UUIDs. Iinkqubo zeshishini ($99/mo) ziquka unikezelo lofolo oluphambili loqhubekeko lweqela olukhawulezayo. Ilungele ukuveliswa kweencwadi zesandi, imixholo yenkqubo, kunye neeprojekthi ezinkulu zesandi.
4.0/5 (8)

Qala Ukusebenzisa i-AI Voice Namhlanje

Dibanisa abavelisi, abaphuhlisi, kunye neenkampani usebenzisa i-TTS.ai

Zama simahla Ngoku Uxwebhu lwe-API