Umbhalo usuka kumazwi Abasebenzi Umculo Ukukhuluma kuMbhalo Ukulungiswa kwezwi Uhlelo lwezwi Imakethe Okuningi

I-Free AI Umbhalo usuka kumazwi

20+ imodeli yomthombo ovulekile 107+ izizwi, 32+ Izilimi. Akukho akhawunti edingekayo.

1K+
abakhiqizi
2K+
izizukulwane
20+
Amamodeli we-AI
107+
imisindo
0/500 amaphawu Ikhululekile
5,000 chars per generation 15,000 amaphawu amahhala No credit card Ukusetshenziswa kwebhizinisi OK
Sign up free
0:00 / 0:00
Layisha phezulu umsindo Isixhumanisi siphele ngehora le-24
Uthanda i-TTS.ai? Xhumana nabangane bakho!
Advanced TTS tool with 20+ models, speed control & file export

Konke okudingayo ngezwi AI

Amathuluzi angama-30+ asebenza ngemodeli ye-AI evulekile

Umbhalo usuka kumazwi

20+ amamodeli we-AI, 100+ amazwi, 30+ ulwimi

Ukukhuluma kuMbhalo

I-Whisper, i-Faster Whisper, SenseVoice. Izilimi ezingu-99

Bhala umsindo

Bhala umsindo nevidiyo ube umbhalo nge-AI

Ukulungiswa kwezwi

Uhlu lwezinhlamvu ezingu-5 zesandi

Izingxoxo zomsindo

Ukukhuluma ne-AI ngezwi elijwayelekile ngesikhathi sangempela

Abasebenzi be-AI

Izwi elinamandla AI lomsebenzi womthengi & okuningi

Umculo we-AI

Dala umculo wokuqala kusuka kumambhalo achazayo

Umshintshi womsindo

Tshintsha isibizo somsindo, inhlanga, isici

Isithuthukisi somsindo

Susa umsindo, thuthukisa ucacile, upscale quality

Ukususa umsindo

Thola ama-vocals noma ama-instrumental kusuka kunoma iyiphi indima

Isihlukanisi somsindo

Yahlula ama-albhamu abe ngama-drums, ama-bass, ama-vocals, nezinye

Ukuhumusha kwezwi

Guqula ulwimi ugcine umsindo womsindo

Ukukhuluma ku-Ukukhuluma

Ukuguqula umsindo, umqondo, ulwimi, uhlobo

Ukususa i-Echo

Susa i-echo negumbi lokuphindaphinda

Iqhosha le-BPM Finder

Thola inkinobho yomculo kanye ne-tempo ngokushesha

Umguquli womsindo

Guqula MP3, WAV, FLAC, OGG, M4A

Umfundi wedokhumende

Bheka ama-athikili, ama-PDF, namadokhumende

Istudio yomsindo

Amabhuku omsindo namapodcast anezinhlayiyana eziningi

Umkhiqizi wepodcast

Ukuxhumana kwe-AI phakathi kwama-host amabili kusuka kunoma iyiphi sihloko

Imiphumela yomsindo

Dala izinguquko zomsindo kusuka kumambhalo abuza

I-batch TTS

Guqula amathegi amakhulu nge-CSV upload

I-Dubbing Studio

I-Dub ividiyo yomsindo yizilimi ezingaphezu kuka-30

Faka i-Widget

Engeza i-TTS kunoma iyiphi iwebhusayithi ngelayini eyodwa

20+ Amamodeli omsindo we-AI

Uhlelo oluphelele kakhulu lwezimo ze-TTS ezivulekile ezikhona kwi-platform eyodwa

KokoroKokoro

Free

Imodeli elula yepharamitha engu-82M ehlinzeka ngolimi olusezingeni le-studio nge-inference esheshayo.

82M parameters Qhubeka kakhulu Izizwi ezichazayo Izilimi eziningi Usizo lokusakaza
Fast · 1.5GB VRAM Zama

PiperPiper

Free

I-neural text to speech system ehamba ngokushesha, etholakala endaweni elungele i-Raspberry Pi nezisetshenziswa ezifakwe ngaphakathi.

CPU-friendly Isebenza ngaphandle kwenethiwekhi Izizwi ezingaphezu kuka-100 Izilimi ezingaphezu kuka-30 Inkxaso ye-SSML
Fast · 0 (CPU only) VRAM Zama

VITSVITS

Free

I-conditional variation autoencoder ene-adversarial learning ye-end-to-end text-to-speech.

Isingeniso esisha I-prosody ejwayelekile Ukubikezela okukhawulelwe Abakhulumayo abaningi
Fast · 1GB VRAM Zama

MeloTTSMeloTTS

Free

Umbhalo osezingeni eliphakeme okhuluma izilimi eziningi osebenza ku-CPU ngesikhathi esincane.

CPU-optimized Izilimi eziningi Isi-Latin Ukukhiqizwa-kulungile Latency ephansi
Fast · 0.5GB (GPU optional) VRAM Zama

BarkBark

Standard

Imodeli yokubhala-kuya-kwesandi esekelwe ku-transformer ekhiqiza amagama acacile, umculo, kanye nemiphumela yomsindo.

Izinhlamvu zomsindo Ukuhlukumeza/ukuphuza Ukukhishwa komculo Abakhuluma ngo-100+ Izilimi eziningi
Slow · 5GB VRAM Zama

Bark SmallBark Small

Standard

Uhlobo oluncane lwe-Bark olunezincazelo ezisheshayo nokusetshenziswa okuphansi kwememori.

Isisindo esincane Ihamba ngokushesha kune-Bark egcwele Ukukhuluma okunengqondo Izilimi eziningi
Medium · 2GB VRAM Zama

CosyVoice 2CosyVoice 2

Standard

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

Ukusakazwa Ukuklona okungenalutho Isilimi esihlukene Ukulawula imizwa I-human-parity
Medium · 4GB VRAM Zama

Dia TTSDia TTS

Standard

Imodeli yokukhiqiza umsindo oningi owenza ukuxhumana okujwayelekile phakathi kwama-speakers.

Isikhulumi esiningi Ukukhiqizwa kwengxoxo Ukuthatha umjikelezo ojwayelekile Isibonakaliso somqondo Amapharamitha we-1.6B
Medium · 4GB VRAM Zama

Parler TTSParler TTS

Standard

Sichaza umsindo ofuna ngesilimi esijwayelekile futhi i-Parler ikhiqiza umsindo olinganayo.

Incazelo yomsindo Ukulawula ulwimi olujwayelekile Ukwakha umsindo osheshayo Akunamagama asethelwe ngaphambili adingekayo
Medium · 4GB VRAM Zama

GLM-TTSGLM-TTS

Standard

Ithola iphutha lophawu oluphansi phakathi kwemodeli ye-TTS yomthombo ovulekile.

Iphutha eliphansi Ukuklona umsindo Ukulandelana kwe-flow I-prosodia ejwayelekile
Medium · 4GB VRAM Zama

IndexTTS-2IndexTTS-2

Standard

I-TTS engekho emthethweni ene-fine-grained emotional control ne-high expressionality.

Ukulawula imizwa I-zero-shot I-emotions vector Ulimi oluchazayo Ukulawula okuncane
Medium · 4GB VRAM Zama

Spark TTSSpark TTS

Standard

Uhlu lwezwi lokuklonya i-TTS nge-emoji elawulwayo nesimo sokukhuluma nge-prompts.

Ukuklona umsindo Ukulawula imizwa Ukulawula isitayela Isekelwe ku-prompt Ukuklona kwesekondi ezingu-5
Medium · 4GB VRAM Zama

GPT-SoVITSGPT-SoVITS

Standard

Uhlu lwezwi lokuklonya TTS oluncane oluphindayo noma yiluphi ulwimi kusuka kumasekondi angama-5 kuphela wesandi.

5-second cloning Umsindo wokungoma Ukufunda okuncane Ikhwalithi ephezulu Isilimi esihlukene
Slow · 6GB VRAM Zama

OrpheusOrpheus

Standard

Imodeli ye-TTS enamandla okuqonda esezingeni lomuntu eqeqeshiwe ngehora le-100K ledatha yokukhuluma.

Umbono womuntu 100K amahora okuqeqeshwa Ukugcizelela okujwayelekile Ukukhuluma okuzwakalayo
Medium · 4GB VRAM Zama

ChatterboxChatterbox

Premium

Uhlelo olusha lokuklonya umsindo olungenalutho olune-emotion control oluvela ku-Resemble AI.

Ukuklona okungenalutho Ukulawula imizwa Ikhwalithi ephezulu Ukudluliswa kwesimo Ukuklona isampula eyodwa
Medium · 4GB VRAM Zama

Tortoise TTSTortoise TTS

Premium

Umbhalo-ku-ukukhuluma okhuluma ngezilimi eziningi obhekene nekhwalithi ngesakhiwo esibuyela emuva.

Ubunjani obuphezulu kakhulu Izwi eliningi DALL-E architecture Ukuklona umsindo Ukubuyela emuva okuzenzakalelayo
Slow · 8GB VRAM Zama

StyleTTS 2StyleTTS 2

Premium

Uhlelo lokuhlela amagama ngokuya ngesimo sengqondo somuntu kanye noqeqesho oluphikisanayo.

Izinga lomuntu Isitayela sokusabalalisa Ukuqeqeshwa kokulwa Ushintsho olujwayelekile Ikhwalithi ephezulu
Medium · 4GB VRAM Zama

OpenVoiceOpenVoice

Premium

Ukuklonya umsindo ngokuzenzakalela ngokulawula okuqinile ngesitayela, inkanuko, nesimo.

Uklonyeliswa okuzenzakalelayo Ukuguqulwa kwezwi Ukulawula imizwa Ukulawula isici Izilimi eziningi
Medium · 4GB VRAM Zama

Qwen3 TTSQwen3 TTS

Standard

I-Alibaba's multilingual TTS nezwi lokuklonya, izizwi ezisetshenzisiwe, kanye nobuciko bezwi kusuka kumbhalo.

Ukuklona umsindo Izizwi ezisetshenzisiwe ezingu-9 Uhlelo lwezwi kusuka kumbhalo Ukulawula imizwa Izilimi
Medium · 7GB VRAM Zama

Sesame CSMSesame CSM

Premium

Imodeli yokukhuluma ekhuluma ngokuzimela ekhiqiza ukuxhumana okujwayelekile ngesikhathi esifanele kanye nemizwa.

Ukuxhumana Isikhathi esijwayelekile Ukushintsha-shintsha Isixhumanisi esingaphambili Amapharamitha we-1B
Slow · 8GB VRAM Zama

KokoroKokoro Free

I-Kokoro iyimodeli ye-text-to-speech eneparameter engu-82 million eyenza kahle ngaphezu kwe-weight class yayo. Nakuba incane kakhulu, ikhiqiza amagama acacile futhi acacile. I-Kokoro isekela izilimi eziningi kufaka phakathi isiNgisi, isiJaphani, isiTshayina, nesiKoreane ngezinhlobonhlobo zamazwi acacile. Isebenza ngokushesha kakhulu — ikhiqiza umsindo osheshayo cishe ngama-100x kunosikhathi sangempela kwi-GPU.

Okungcono kakhulu: Ikhwalithi ephezulu ye-TTS enesikhathi sokuphuma esincane, izisebenziso zokusakaza

Zama mahhala

PiperPiper Free

I-Piper iyinjini elula yokubhala-ukukhuluma ethuthukiswe yi-Rhasspy esebenzisa i-VITS ne-larynx architectures. Isebenza ngokuphelele ku-CPU, iyenza ibe ngcono kakhulu kumadivayisi e-edge, ukuphathwa kwekhaya, namathuluzi adinga i-TTS engenayo. Ngezwi elingaphezu kuka-100 lidlula ulwimi olungaphezu kuka-30, i-Piper inikeza ukukhuluma okubukekayo ngokuzenzakalela ngejubane lesikhathi sangempela ngisho ne-Raspberry Pi 4.

Okungcono kakhulu: Ukubukeka okukhawulelwe, ukufinyeleleka, kanye nezisebenziso ezifakwe ngaphakathi

Zama mahhala

VITSVITS Free

VITS (Izibalo ezishintshayo ezifunda ngokuphikisanayo ukuqala ukubhala-ukukhuluma-ukuphela-ku-kuphela) yindlela ye-TTS elinganayo ekugcineni-ku-kuphela ekhiqiza umsindo ozwakalayo ojwayelekile kunalezo ezingemuva-ezimbili. Isebenzisa izibalo ezishintshayo ezithuthukisiwe ngokuhamba okujwayelekile kanye nenqubo yokuqeqeshwa okuphikisanayo, ethola ukukhula okuphawulekayo ekungavamile.

Okungcono kakhulu: Umbhalo-ku-ukukhuluma okusetshenziswa kakhulu nge-prosody ejwayelekile

Zama mahhala

MeloTTSMeloTTS Free

MeloTTS ngu MyShell.ai yi-TTS library eminingi ye-languages exhasa isiNgisi (i-American, i-British, i-Indian, i-Australian), isiShayina, isiJalimane, isiKorean. Ishesha kakhulu, isebenza umbhalo ngejubane elifanayo nesikhathi sangempela kwi-CPU kuphela. MeloTTS isetshenziselwa ukusetshenziswa kokukhiqizwa futhi ixhasa i-CPU ne-GPU inference.

Okungcono kakhulu: Izisebenziso zokukhiqiza ezidinga i-TTS esheshayo, enezilimi eziningi

Zama mahhala

BarkBark Standard

Imodeli yokubhala-kuya-kwesandi esekelwe ku-transformer ekhiqiza amagama acacile, umculo, kanye nemiphumela yomsindo.

Umthuthukisi: Suno · Ilayisense: MIT

Zama

Bark SmallBark Small Standard

Uhlobo oluncane lwe-Bark olunezincazelo ezisheshayo nokusetshenziswa okuphansi kwememori.

Umthuthukisi: Suno · Ilayisense: MIT

Zama

CosyVoice 2CosyVoice 2 Standard

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

Umthuthukisi: Alibaba (Tongyi Lab) · Ilayisense: Apache 2.0

Zama

Dia TTSDia TTS Standard

Imodeli yokukhiqiza umsindo oningi owenza ukuxhumana okujwayelekile phakathi kwama-speakers.

Umthuthukisi: Nari Labs · Ilayisense: Apache 2.0

Zama

Parler TTSParler TTS Standard

Sichaza umsindo ofuna ngesilimi esijwayelekile futhi i-Parler ikhiqiza umsindo olinganayo.

Umthuthukisi: Hugging Face · Ilayisense: Apache 2.0

Zama

GLM-TTSGLM-TTS Standard

Ithola iphutha lophawu oluphansi phakathi kwemodeli ye-TTS yomthombo ovulekile.

Umthuthukisi: Zhipu AI · Ilayisense: GLM-4 License

Zama

IndexTTS-2IndexTTS-2 Standard

I-TTS engekho emthethweni ene-fine-grained emotional control ne-high expressionality.

Umthuthukisi: Index Team · Ilayisense: Bilibili Model License

Zama

Spark TTSSpark TTS Standard

Uhlu lwezwi lokuklonya i-TTS nge-emoji elawulwayo nesimo sokukhuluma nge-prompts.

Umthuthukisi: SparkAudio · Ilayisense: CC BY-NC-SA 4.0

Zama

GPT-SoVITSGPT-SoVITS Standard

Uhlu lwezwi lokuklonya TTS oluncane oluphindayo noma yiluphi ulwimi kusuka kumasekondi angama-5 kuphela wesandi.

Umthuthukisi: RVC-Boss · Ilayisense: MIT

Zama

OrpheusOrpheus Standard

Imodeli ye-TTS enamandla okuqonda esezingeni lomuntu eqeqeshiwe ngehora le-100K ledatha yokukhuluma.

Umthuthukisi: Canopy Labs · Ilayisense: Llama 3.2 Community

Zama

Qwen3 TTSQwen3 TTS Standard

I-Alibaba's multilingual TTS nezwi lokuklonya, izizwi ezisetshenzisiwe, kanye nobuciko bezwi kusuka kumbhalo.

Umthuthukisi: Alibaba (Qwen) · Ilayisense: Apache 2.0

Zama

ChatterboxChatterbox Premium

Uhlelo olusha lokuklonya umsindo olungenalutho olune-emotion control oluvela ku-Resemble AI.

Ubunjani:

Zama

Tortoise TTSTortoise TTS Premium

Umbhalo-ku-ukukhuluma okhuluma ngezilimi eziningi obhekene nekhwalithi ngesakhiwo esibuyela emuva.

Ubunjani:

Zama

StyleTTS 2StyleTTS 2 Premium

Uhlelo lokuhlela amagama ngokuya ngesimo sengqondo somuntu kanye noqeqesho oluphikisanayo.

Ubunjani:

Zama

OpenVoiceOpenVoice Premium

Ukuklonya umsindo ngokuzenzakalela ngokulawula okuqinile ngesitayela, inkanuko, nesimo.

Ubunjani:

Zama

Sesame CSMSesame CSM Premium

Imodeli yokukhuluma ekhuluma ngokuzimela ekhiqiza ukuxhumana okujwayelekile ngesikhathi esifanele kanye nemizwa.

Ubunjani:

Zama

CosyVoice 2CosyVoice 2

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

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

Umsindo

GLM-TTSGLM-TTS

Ithola iphutha lophawu oluphansi phakathi kwemodeli ye-TTS yomthombo ovulekile.

Izilimi: en, zh

Umsindo

IndexTTS-2IndexTTS-2

I-TTS engekho emthethweni ene-fine-grained emotional control ne-high expressionality.

Izilimi: en, zh

Umsindo

Spark TTSSpark TTS

Uhlu lwezwi lokuklonya i-TTS nge-emoji elawulwayo nesimo sokukhuluma nge-prompts.

Izilimi: en, zh

Umsindo

GPT-SoVITSGPT-SoVITS

Uhlu lwezwi lokuklonya TTS oluncane oluphindayo noma yiluphi ulwimi kusuka kumasekondi angama-5 kuphela wesandi.

Izilimi: en, zh, ja, ko

Umsindo

ChatterboxChatterbox

Uhlelo olusha lokuklonya umsindo olungenalutho olune-emotion control oluvela ku-Resemble AI.

Izilimi: en

Umsindo

Tortoise TTSTortoise TTS

Umbhalo-ku-ukukhuluma okhuluma ngezilimi eziningi obhekene nekhwalithi ngesakhiwo esibuyela emuva.

Izilimi: en

Umsindo

OpenVoiceOpenVoice

Ukuklonya umsindo ngokuzenzakalela ngokulawula okuqinile ngesitayela, inkanuko, nesimo.

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

Umsindo

Qwen3 TTSQwen3 TTS

I-Alibaba's multilingual TTS nezwi lokuklonya, izizwi ezisetshenzisiwe, kanye nobuciko bezwi kusuka kumbhalo.

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

Umsindo

Umthuthukisi-kuqala API

I-REST API ehambisana ne-OpenAI. Ingxenye eyodwa, amamodeli angama-22+ Ukusakazwa kwengxoxo yesikhathi sangempela.

  • Ifomethi ehambisana ne-OpenAI
  • Ukusakazwa kwe-TTS kwezinhlelo zokusebenza zesikhathi sangempela
  • Uhlelo lwe-batch lwemisebenzi enkulu
  • Ulwaziso lwe-Webhook
Bona amadokhumende we-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")

Intengo elula, ecacile

Qalisa ngokukhululekileyo. Ukukala njengoba ukhula.

Ikhululekile

$0

15,000 characters

  • Kokoro, Piper, VITS, MeloTTS
  • Iphutha lophawu lwe-500
  • 3 gen/ihora (akukho akhawunti)
Ubhalise

Isiqalisi

$9/ihora

500,000 characters/month

  • Zonke imodeli ezingu-22+
  • 100,000 chars per generation
  • Ukulungiswa kwezwi
Qala
Okuthandwa kakhulu

I-Pro

$29/ihora

2,000,000 characters/month

  • Konke ku-Starter
  • Ukungena kwe-API
  • Ukulungiswa kokuqala
Thola i-Pro

Ibhizinisi

$99/ihora

10,000,000 characters/month

  • Konke ku-Pro
  • I-bulk API
  • Ifolokhwe yesinqumo
Thola umsebenzi

Bona zonke izilungiselelo kufaka phakathi izilungiselelo zophawu →

Imibuzo ebuzwa kaningi

I-TTS.ai iyipulatifomu yezwi le-AI ebanzi kakhulu, enikeza amamodeli we-text-to-speech angama-22 +, ukucloning kwezwi, ukuxoxa kwe-text-to-text, namathuluzi e-audio. Onke amamodeli avulekile ngaphandle kwe-vendor lock-in.

Yebo! TTS.ai inikeza umbhalo-ku-ukukhuluma mahhala nge-Kokoro, Piper, VITS, ne-MeloTTS models. Akukho akhawunti edingekayo. Bhala ukuze uthole ama-15,000 ama-characters mahhala futhi ufinyelele kuzo zonke imodeli. Ama-plans akhokhelwayo aqala ku- $ 9 / ngenyanga.

Ukukhawulela, sebenzisa iKokoro noma iPiper. Ukwenza kahle, sebenzisa iCosyVoice 2 noma iStyleTTS 2. Ukuklona umsindo, sebenzisa iChatterbox noma iGPT-SoVITS. Ukwenza ingxoxo, sebenzisa iDia TTS. Zama amamodeli amaningi ku mbhalo owodwa ukuwalinganisa.

Yebo. I-OpenAI-ehambisana ne-REST API ye-TTS, i-STT, ukucloning kwezwi, namathuluzi e-audio. Itholakala ku-Pro ($ 29 / mo) ne-Enterprise ($ 99 / mo) amaphrojekthi. Bona idatabase ku-tts.ai / api /.

Ikhwalithi yomsindo ihluka ngokwemodeli. Amamodeli aphezulu njenge-CosyVoice 2, StyleTTS 2, ne-Chatterbox akhiqiza umsindo osezingeni elifanayo nomuntu nge-intonation ne-emotions ezijwayelekile. Amamodeli amahhala njenge-Kokoro anikeza ikhwalithi engcono kakhulu yezimo eziningi zokusetshenziswa.

I-TTS.ai ixhasa izilimi ezingaphezu kuka-30 ngaphesheya kwelayibrari yayo yemodeli. IsiNgisi sinesixhaso semodeli esibanzi kakhulu, kodwa amamodeli afana ne-CosyVoice 2 afaka ama-Chinese, ama-Japanese, nama-Korean; i-GPT-SoVITS iphatha ama-Chinese, ama-Japanese, ama-Korean, nama-English; futhi i-MeloTTS ixhasa ama-English, ama-Spanish, ama-French, ama-Chinese, ama-Japanese, nama-Korean.

Yebo. Zonke izisebenziso zikhona kumaseva ethu akhethekile we-GPU. Asigcinanga umbhalo wakho ongeniswe noma umsindo okhiqizwe ngemuva kokuthunyelwa. Izinhlamvu zomsindo ezilayishwe phezulu zokuklonya zisetshenziswa kuphela kwisiqephu samanje futhi azigcinwanga. Asikwazi ukuhlukanisa idatha yakho namanye amaqembu noma ukusebenzisa imodeli yokuqeqesha.

Yebo. Zonke izisindo ezikhiqizwa ku-TTS.ai zikhona kuwe ukuze uzisebenzise ngokuhweba, kufaka phakathi i-YouTube video, i-podcast, ama-audiobooks, ama-apps, izikhangiso, nama-products. Amamodeli ethu avulekile ngaphansi kwelayisense elivumelayo (MIT, Apache 2.0). Akukho lungelo noma ukuphawula okudingekayo.

I-TTS.ai ikhiqiza umsindo ngefomethi ye-WAV ngokuzenzakalela ngekhwalithi ephezulu. Ungaguqula ube yi-MP3, FLAC, OGG, noma i-M4A usebenzisa ithuluzi lethu elimahhala le-Audio Converter. I-API ixhasa ukucacisa ifomethi yakho yokukhishwa okuthandekayo ngqo kwisicelo.

Layisha phezulu isampula lesandi esincane (esincane njengemizuzu emihlanu) sezwi ofuna ukuliklonela, bese ubhala noma iyiphi incwadi ukudala ukukhuluma kulolu zwi. Amamodeli afana ne-Chatterbox, GPT-SoVITS, ne-CosyVoice 2 axhasa ukuklonela kwezwi. Uzwi oluklonelwe luthatha into, umbala, nesimo sokukhuluma.

Amamodeli amahhala (iKokoro, iPiper, iVITS, iMeloTTS) adinga i-akhawunti futhi abiza ama-characters ayi-zero. Amamodeli ajwayelekile (ama-characters angama-2,000/1K) afaka iBark, iCosyVoice 2, iF5-TTS, neDia. Amamodeli aphezulu (ama-characters angama-4,000/1K) afaka iOpenVoice, iChatterbox, iStyleTTS 2, neTortoise. Amamodeli akhokhelwayo ngokuvamile anikeza umgangatho ophezulu, amazwi amaningi, kanye nezici ezingeziwe ezifana nokuklonyelwe kwezwi.

Yebo. I-API ixhasa ukucutshungulwa kwe-batch ukuguqula amavolumu amakhulu we-text ku-speech. Sebenzisa izicelo eziningi bese uthola izimpendulo nge-asynchronously usebenzisa i-job UUIDs. Izinhlelo ze-Enterprise ($99/mo) zifaka phakathi ukufinyelela kwe-queue yokuqala ukucubungula okusheshayo kwe-batch. Ilungele ukukhiqizwa kwe-audiobook, okuqukethwe kwe-course, namaphrojekthi amakhulu we-voiceover.
4.0/5 (8)

Qala ukusebenzisa umsindo we-AI namhlanje

Xhumana nabakhiqizi, abathuthukisi, namabhizinisi asebenzisa i-TTS.ai

Zama mahhala manje Amadokhumende we-API