I-Shell

Premium

Ukukhuluma uma ukhuluma ngemizwa nemiphumela yomsindo

Zama mahhala I-Shell Hlola izizwi (130)
Slow Isivinini
Very Good Ubunjani
Akukho Ukuklonya
13 Izilimi

Ngo I-Shell

-text audio. Bark is a

Izici ezibalulekile

Ukubonisa umqondo

Yenza ukukhuluma ngokuthanda, ukuxosha, ukuxosha, nokuzizwa okungokoqobo.

Amaphawu emizwa

Sebenzisa [ukuthanda], [ukuzizwa], CAPS ukuphawula, no... ukukhathazeka.

Izilimi eziningi

Insiza 13+ izilimi ezinama-accents ajwayelekile kanye nokukhuluma.

Umculo & Izinhlekelele

Ingadala umculo olula kanye nezingcingo zendawo.

Izilungiselelo zomsindo

Izizwi eziningi eziqeqeshiwe zomsindo ezinamamodeli ahlukene.

Umsuka ovulekile

I-MIT ilayisensewe ngelungelo lokusebenziseka okuphelele.

Sebenzisa izimo

Ibhokisi lemiyalezo Okuqukethwe okuhlanganisiwe Ukukhuluma incwadi yomsindo Umdlalo wezwi Iphrojekthi eyenziwe Ama-Asistents achazayo

I-Shell Voices

View All 130
Bark Chinese Speaker 0
ZH
Bark Chinese Speaker 1
ZH
Bark Chinese Speaker 2
ZH
Bark Chinese Speaker 3
ZH
Bark Chinese Speaker 4
ZH
Bark Chinese Speaker 5
ZH
Bark Chinese Speaker 6
ZH
Bark Chinese Speaker 7
ZH
Bark Chinese Speaker 8
ZH
Bark Chinese Speaker 9
ZH
Bark English Speaker 0
EN
Bark English Speaker 1
EN

Indlela yokusetshenziswa I-Shell

  1. 1

    Ubhalise mahhala futhi uvule idemo

    Dala i-akhawunti emahhala ye-TextToSpeechAI ukuze uqinisekise ama-credits akho wokuqamba, noma sebenzisa idemo engabhaliswanga ukuzama i-Bark ngokushesha. Ama-credits amahhala afinyelela ukuletha ama-clip we-Bark ambalwa abonisayo ngaphambi kokufaka ushintsho.

  2. 2

    Khetha umsindo we-Bark

    Vula i-library yomsindo bese ukhetha umsindo we-Bark osethelwe ngaphambili olingana ne-tone oyifunayo. Amazwi we-Bark aphawulwa njenge-premium level (ama-credits angama-25 ngamagama angama-1000) futhi alungele ukucabanga, ukuchaza amagama.

  3. 3

    Faka umbhalo osebenzisa izibonakalisi zemizwa

    Bhala isikripthi sakho bese ufaka izibonisi zesifiso se-Bark ezingaphakathi: [uthando] ukulalela, [uthando] ukulalela, [uthando] ukulalela,... ukuphumula, ne-CAPS ukuphawula. Umzekelo: "Oh wow! [uthando] Le yindlela emangalisayo... Ngingakholwa!"

  4. 4

    Dala umsindo

    Chofoza i-Generate futhi i-Bark inikeza umbhalo wakho umsindo ozwakalayo, iguqula wonke umqondisi ube ngumsindo ohambisanayo. Ukukhiqizwa kushesha kunezinjini ezincane ngenxa yemodeli yokuguqula i-Bark, ngakho vumela imizuzwana embalwa engeziwe ngamazwi.

  5. 5

    Layisha phezulu noma sebenzisa i-API

    Phinda ubheke imiphumela, bese uyilanda njenge MP3, WAV, noma OGG. Ukwenza i-Bark isebenze ngokuzenzakalela kwi-app yakho, thinta i-TextToSpeechAI API ngezwi le-Bark kanye nombhalo ofanayo ogcwele umgcini-mbhalo ukuze ubuyisele umsindo obonisayo.

I-Shell 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": "Oh wow! [ukuthanda] Le nto imangalisa... Ngithanda kakhulu ukuthi lokhu kubukeka kanjani!",
    "voice": "bark-zh_0"
  }'
Funda amadokhumende we-API Thola isithonjana sakho se-API

Imibuzo ebuzwa kaningi

I-Bark iyimodeli yokubhala-kuya-kwesandi esekelwe ekushintsheni eyenziwe yi-Suno. Ngokungafani nezinhlelo ze-TTS ezijwayelekile, i-Bark ikhiqiza amagama azwakalayo kakhulu ngemizwa ejwayelekile, ukulalela, ukuphefumula, nezinye izizwi ezingasho lutho. Ingakhiqiza nama-music nama-sound effects.

Yebo, i-Bark ivela kumthombo ovulekile ngaphansi kwelayisense ye-MIT, evumela ukusetshenziswa okumahhala kokuthengiswayo. Ku-TextToSpeechAI, sikhokhisa ama-credits angama-25 ngamagama angama-1000 ngenxa yama-GPU abalulekile adingekayo ukuthuthukiswa.

I-Bark isekela izilimi ezingaphezu kuka-13 kufaka phakathi isiNgisi, isiJalimane, isiSpanishi, isiFulentshi, isiHindi, isiItalian, isiJapane, isiKorea, isiPoland, isiPutukezi, isiRussia, isiTurkey, nesiChinese. Isilimi ngasinye sinezinhlamvu ezijwayelekile nezilimi ezihlukile.

I-Bark ihamba kancane kunezinye i-TTS engines ngenxa yobuciko bayo be-autoregressive transformer. Igama elijwayelekile lithatha amasekondi angama-5-15 ukukhishwa ku-GPU. Ushintsho luvela ngokucacile futhi luvela ngokujwayelekile.

I-Bark inikeza kuphela ukuklonywa kwezwi okulinganiselwe nge-"semantic prompts" ne-speaker preset, ngakho-ke ayikwazi ukuklonya ngokuthembekile izwi elifunayo kusuka kusampula. Uma ukuklonywa kwezwi okugcwele kuyinhloso yakho, sebenzisa i-F5-TTS, StyleTTS2, OpenVoice, noma i-Tortoise endaweni, zonke zikhona ku-TextToSpeechAI.

I-Bark ifunda izibonisi ezingaphakathi kwelayini ezibekwe ngokuqondile embhalweni wakho futhi iguqula zibe izingcingo ezihambisanayo. Sebenzisa [uthando] ukulalela, [uthando] ukulalela, [uthando] ukulalela,... ukukhathazeka noma ukuphumula, ne-CAPS ukuphawula. Umzekelo: "Oh wow! [uthando] Le yindlela emangalisayo... Ngingakholwa!"

Ngaphandle kokukhuluma okujwayelekile, i-Bark ingadala amazwana angasho lutho njengenhlanhla, ukumamatheka, ukumamatheka, ukucisha ikhanda, nokumamatheka, kanye nemiculo elula kanye nemiphumela yendawo. Lezi ziqhutshwa ngama-markers njenge [inhlanhla], [ukumamatheka], kanye [ukumamatheka] efakwe kuthebhu, okuzokwenza i-Bark ibonakale izwakala kakhulu kune-TTS ejwayelekile.

I-Bark ikhiqiza umsindo osezingeni eliphakeme kakhulu nge-expression ejwayelekile efana nokukhuluma umuntu ngezinto ezithinta imizwa. I-24kHz output izwakala isebenza kahle, kepha ukukhuluma okulula kungaphansi kwe-StyleTTS2.

I-Bark idinga i-8-12GB ye-VRAM ngokuya ngesayizi yemodeli. Imodeli ephelele idinga i-~12GB, kanti izinketho ezincane zisebenza nge-8GB. Ukuqonda kwe-CPU kunzima kakhulu futhi akudingeki.

Yebo, i-Bark i-MIT licensed, evumela ukusetshenziswa okuhweba okungagcini ngaphandle kwezindleko zokulingisa. Ungasebenzisa i-Bark kuma-products, amasevisi, nama-applications ngokukhululekile. Ku-TextToSpeechAI ungazama i-Bark ngaphandle kokusebenzisa ama-credits akho okubhalisa ngaphambi kokhokha okuningi.

I-Bark ihamba phambili ekubhaleni okucacile kwe-single-speaker nge-emotions markers njenge [ukuthanda] ne [ubuhlungu], ngenkathi i-Dia ifakwe ukuxhumana kwe-multi-speaker nge-[S1]/[S2] turns kanye ne-nonverbal cues. Khetha i-Bark ukukhuluma okujabulisayo nokukhuluma komuntu, ne-Dia ukukhuluma okubuyela emuva nokuhamba phambili. Zonke zikhona ku-TextToSpeechAI.

I-Bark ihlukile ngokukwazi kwayo ukudala amagama acacile abonisa imizwa nezingcingo ezingasho lutho. Ihamba kancane kunezinye izinjini kodwa ikhiqiza izimpendulo ezifana nomuntu ezingaphezu kwalezi eziqukethwe ezizimele. Ukwenza isizinda esisheshayo, sebenzisa i-Piper. Ukwenza isizinda esifanayo, sebenzisa i-F5-TTS noma i-OpenVoice.

Technical Specs

  • Generation Speed Slow
  • Output Quality Very Good
  • Voice Cloning Not Supported
  • Languages 13
  • GPU VRAM 8-12GB
  • Credits/1000 chars 25

Try I-Shell Now

Generate your first audio free. No credit card required.

Start Free

Other TTS Engines

Ibhokisi lokuxoxa
Fast | Very Good
Isibizo esimnandi 2
Fast | Very Good
Imini
Medium | Excellent
View All 9 Engines