Powered by Chatterbox, Kokoro & Whisper

I-Ultra
Ukuthetha-thethana
Inkqubo ye-AI

Ukwenza ikopi yelizwi ngasekuqaleni, uguqule umbhalo ube ngumbhalo, uguqule umsindo ube ngumsindo, uguqule ilizwi libe elinye, zonke ngeemodeli ze-state-of-the-art open-source.

Akukho khadi letyala lifunekayo 10 credits free Ufikelelo lwe-API luquka
Imboniselo YokushicilelaComment
Amagama afanayo, ilizwi elahlukileyo

Ukufunda

Igama lefayile le CVS:

I-Agile

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

Yonke into oyifunayo kwi Voice AI

Izixhobo ezipheleleyo zesicelo sokuthetha

Indlela esebenza ngayo

Iindlela ezintathu ezilula zokutshintsha kwelizwi

1

Layisha phezulu ifayile ye PDF

Layisha phezulu isampuli yelizwi (10-30 imizuzwana) okanye urekhode ngqo kwibrowser yakho. Sixhasa i MP3, WAV, FLAC, nezinye.

2

Inkqubo ye-AI

Iiseva zethu ze-GPU ziqhubekekisa umsindo wakho ngeemodeli ezivulekileyo ezinjengeChatterbox, Kokoro, Whisper kunye neOpenVoice.

3

Layisha ezantsi okanye i-stream

Fumana iziphumo zakho ngokuzenzekelayo. Layisha ezantsi kwifomati ezininzi okanye udlulise ngokuthe ngqo nge-API yethu yexesha elibonakalayo.

Isebenza ngeemodeli ezilungileyo ezivulekileyo

Sisebenzisa kwaye sixhasa uvavanyo oluphambili lwe-AI

OpenVoice v2

Injini yokuguqulela ilizwi elikhethiweyo kunye nemodeli yokuklonya

  • Uguqulelo lombala we-Tone
  • Igcina amagama nexesha
  • Ilayisensi ye-MIT
Chatterbox

Ukuphinda usebenzise ilizwi, ukuphinda usebenzise ilizwi kunye nokuguqulwa kwelizwi lethu elimiselweyo (Chatterbox VC)

  • Uklonelo lwe-zero-shot
  • Iilwimi ezili-23
  • Ukusetyenziswa kwentengiso kuvunyelwe
Kokoro 82M

Fast, natural text to speech

  • 54 voices
  • 8 languages
  • Apache 2.0 licence
Whisper Large v3

Uguqulelo kolunye ulwimi

  • Ubhaqo oluzenzekelayo lweelwimi
  • Iilwimi ezili-100
  • Igama lexesha le-timestamp

Ifakwe kwimeko yokusetyenziswa nganye

Ukusuka ekudaleni imixholo ukuya kusetyenziso lweshishini

Ukwenza imixholo

Dala iingoma zevidiyo, iipodcasts, iincwadi zesandi, kunye nomxholo wemidiya yoluntu ngokuthe nkqo.

Inkonzo yomthengi

Yakha abancedisi belizwi abathetha ngokuqhelekileyo kwaye bakwazi ukuphatha unxibelelwano olunzima.

Ubeko lwephepha

I-Dub iqulethe imixholo kwiilwimi ezininzi ngelixa igcina ilizwi lomthumeli ophambili.

Ufikelelo

Inika ilizwi kubo bonke abalahlekileyo. Yenza uthungelwano lokuthetha oluzimeleyo.

Imidlalo & Imidlalo

Dala ilizwi elinamandla le-NPC, unxibelelwano lophawu, kunye neengcinga ezingqukuva.

Imfundo

Yenza imixholo yokufunda enomdla equlethe ukuthetha okunje

Umbhekisi phambili

Yenza nge-API yethu

I-REST API elula yesixhobo ngasinye: thumela isicelo ngesitshixo sakho se-API, emva koko ujonge iziphumo.

  • I-RESTful API

    Iindawo ezisiphelo ze-HTTP ezilula kuzo zonke izinto eziluncedo

  • Ii-Webhooks

    Imisebenzi yokuguqulela inga POST iziphumo zayo kwi URL yakho xa igqityiwe

Bonisa i-API Docs
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"}

Ixabiso elilula, elicacileyo

Qala ngokukhululekileyo, i-scale njengoko ukhula

Iinketho ze-GNU/Linux
$0/iinyanga
  • 10 credits/month
  • Zonke iisixhobo ziquka
  • Ufikelelo lwe-API
Qala
Isiqalisi
$9/iinyanga
  • 100 credits / Inyanga
  • Zonke iisixhobo ziquka
  • Ufikelelo lwe-API
Khetha Isiqalisi
I-Pro
$29/iinyanga
  • 500 credits / Inyanga
  • Zonke iisixhobo ziquka
  • Ufikelelo lwe-API
Khetha i-Pro
I-Enterprise
$99/iinyanga
  • 2000 credits / Inyanga
  • Zonke iisixhobo ziquka
  • Ufikelelo lwe-API
Khetha iNkqubo

Imibuzo ebuzwa rhoqo

Zonke izinto ofuna ukuziqonda malunga ne-Speech to Speech AI

Ukufana kwelizwi lisebenzisa i-AI ukuvavanya isampula yelizwi lomntu (i-10-30 imizuzwana) kwaye livelise imodeli yedijithali enokuthi ithethe nayiphi na incwadi kulo mbhalo. Inkqubo yethu isebenzisa imodeli ye-state-of-the-art njengeChatterbox neOpenVoice ukufezekisa iziphumo ezithethayo ngokuqhelekileyo.

Sixhasa isiNgesi, isiSpanish, isiFrentshi, isiJamani, isiTshayina, isiJaphani, isiKorea, isiArabhu, isiPutukezi, isiRashiya, isiTaliyani, isiDutch, isiPolish kunye nezinye ezininzi. Iimodeli zethu eziziilwimi ezininzi zingakhupha ilizwi kwaye zivelise ukuthetha kulo naluphi na lwalawo ulwimi.

Ewe! Sinika umphakamo okhululekileyo oneempawu ezili-10 ngenyanga, oquka ukufikelela okupheleleyo kuzo zonke izinto eziluncedo: ukuclonelwa kwelizwi, umbhalo-ukuthetha, umbhalo-ukuthetha, kunye nezinye izinto. Akukho khadi letyala elifunekayo ukubhalisa. Nciphisa ixesha naliphi na ukuze ufumane ii-credits ezininzi.

Sisebenzisa imodeli ye-open-source: iChatterbox ne OpenVoice v2 yokuhlalutya ukuthetha, iKokoro, iPiper, iVITS ne Bark yokubhaliweyo ukuya kukuthetha, iWhisper Large v3 yokushicilela, iChatterbox VC yokutshintsha ukuthetha (iOpenVoice v2 iyaqhubeka ikhethiwe), iChatterbox yelizwi elihlalutyekayo kwi-AI dubbing kunye neQwen2.5 yempendulo kwincoko yokuthetha. Zonke ziqhuba kwiseva yethu ye-GPU.

Ngokuqinisekileyo! Zonke iinkqubo zethu ezihlawulwayo ziquka ilungelo lokusetyenziswa korhwebo. Ungasebenzisa umsindo owenziweyo kwiividiyo, iipodcasts, iiapps, imidlalo, uncedo lomthengi, nakweyiphi na enye inkqubo yorhwebo. Qinisekisa kuphela ukuba unelungelo lokukrola nayiphi na imiyalezo oyisebenzisayo.

Incoko yelizwi likuvumela ukuba uthethe nomncedisi we AI. Umkhangeli zincwadi wakho uguqula uthetha kwakho kumbhalo, i Qwen2.5 (imodeli yegama elivulekileyo eliqhuba kwi GPU yethu) ibhala impendulo kwi persona okhethayo, kwaye ulizwi le Piper lifunda ngokucacileyo. Isebenza ngokujikeleza ngokujikeleza, ngesiNgesi.

Iinkqubo zekhompyutha ezisebenzisa umyalezo we-imeyili

Dibanisa amawaka ababhekisi phambili nabavelisi besebenzisa i Speech to Speech AI.