Tọghata ngwe ka ọsụsọ Ndịna AI Ụdị Ogo ka ngwe Klọnsị ụda Nhazi ụda Marketplace Oge
VibeVoice

Speaker 1

Dìfọ́ọ̀ltụ̀ English Neutral VibeVoice

Speaker 1 bụ olu neutral AI nke ejiri ike site na móòdù ngwe-ka-asụsụ VibeVoice. Asụsụ standard-tier a na-ekwu English ma na-enye nsụgharị okwu STUDIO-ọdịmma. Na n'ụdị nrụpụta ọsọ nke N'ime nkeji nakwa nkwalite nke 5/5, Speaker 1 dịkwa mma maka podcasts, dialogues, long-form narration, multi-speaker content. Ékèkọ́rá VibeVoice engine site na Microsoft under the MIT license, na-eme ka ọ bụrụ nke dị mma maka ojiji azụmahịa. Nhazi ndị dị mkpa gụnyere: multi-speaker, long-form (90 min), podcast generation, dialogue, low latency.

Enweghị ụghasị
Nwalee ụda a Ha niile VibeVoice Ụda

VibeVoiceNdesịta ozi model

Móòdù VibeVoice
Debanye aha Microsoft
Nhazi
Nhazi Nnọọ
Ikikere MIT
Ọrụ Enweghị ike ịhụ ya
Ụdị Nhazi (2x akara)
Paramita 1.5B
Nhazi LLM + DAC
Ndesịta ozi ndịna 100000 awa
Ụbọchị 2025

Ọrụ kacha mma maka Speaker 1

Usoroiheomume a na-atụ aro site na ụda a

Agụgụala na ndezi

Jiri Speaker 1 ka ịkọwapụta ihenhọrọ nke ogologo-fomu na n'ụzọ na-ezighị ezi nakwa nkọwa.

Vidéọ̀wù

Tinye nkọwa profaịlụ na vidiyo YouTube, mgbasaozi, na ihenhọrọ mgbasaozi mmekọrịta.

Usoroiheomume na ntọala

Nhazi nke n'ụzọ nkịtị na-eme ka ụda a dị mma maka usoroiheomume oge-ọdịnihu, ndị na-agụ ihuenyo, nakwa ngwaọrụ nlebara anya.

Podcasts na mbipụta

Ogo nke ọma redio maka podcasts, nakwa maka mgbasaozi profaịlụ.

Ndị ọzọ VibeVoice Ụda

Ụda ndị ọzọ site na móòdù TTS ahụ

Speaker 1 (Chinese)

Chinese Neutral

Speaker 2

English Neutral

Speaker 2 (Chinese)

Chinese Neutral

Speaker 3

English Neutral

Speaker 4

English Neutral
Gosi _Ha niile VibeVoice Ụda

Ajụjụ ndị a na-ajụkarị

VibeVoice site na Microsoft na-abịa na ụdị abụọ: ụdị 1.5B maka ihe nchọgharị ogologo oge (na-eru minit 90, ndị na-ekwu okwu 4) na ụdị Realtime 0.5B maka ịkpọ egwu na ~ 200ms mbụ ụda latency. Ụdị 1.5B na-egosipụta na podcasts na ụda akwụkwọ na ụda okwu na-egosipụta n'oge ogologo oge. Ntụziaka: Microsoft wepụrụ TTS koodu site na repository ma mepụta ụda gụnyere nkwenye AI dị egwu.

VibeVoice a rụpụtara site na Microsoft ma pụta n'okpuru MIT (n'ihi nchọpụta) ikike, nke na-enye ikike iji ọrụ ọha na eze nke ụda a rụpụtara.

VibeVoice na-akwado asụsụ 1: English.

VibeVoice bụ na Premium tier - 4 credits kwa 1,000 characters. I nwere ike ịhụ n'ihu ọbụla VibeVoice ụda n'efu tupu ịmepụta ụda zuru ezu.

VibeVoice nwere ọsọ mbipụta nke dị ala. Mbipụta na-ewe sekọnd ole na ole dabere na ogologo ngwe ahụ.

VibeVoice a raara 5/5 maka ogo ụda na TTS.ai. O na-enye studio-grade, okwu dị ka mmadụ.

Ee, VibeVoice na-eji setịpụrụ ụda ndị ahụ. Maka ụda ndị ahụ, jiri móòdù dị ka CosyVoice 2, GPT-SoVITS, mọọbụ Chatterbox.

Ya, VibeVoice bụ nke a na-atụ aro maka podcasts, audiobooks, ogologo-ụdị multi-speaker ọdịnaya. Multi-speaker ya, ruo 90 min, podcast mmepe ikikembanye mee ya a dị mma nhọrọ maka a iji ihe omume.

Ee, VibeVoice ejirila ya n'okpuru MIT (n'ihi nnyocha-ọbụla), nke na-enye ohere iji ya maka ọrụ azụmahịa. A ga-eji ụda a haziri na VibeVoice jiri ya rụọ ọrụ na vidio, podcasts, ngwaike, egwuregwu, nakwa ihe ọbụla ọzọ maka ọrụ azụmahịa.

Ee, ụda niile na TTS.ai na-eji ụdị ndị a na-enye ikike n'ụzọ azụmahịa (MIT, Apache 2.0). Ọdịdị a haziri bụ nke gị iji jiri ya na vidio, podcasts, ngwa, egwuregwu, nakwa usoroiheomume ọbụla ọzọ na-enye ikike n'ụzọ azụmahịa.

Ziga arịrịọ POST na /api/v1/tts/ na aha móòdù nakwa ID ụda. Gụọ ibe anyị nke Dọkumenti API maka ihenhọrọ koodị na Python, JavaScript, Go, nakwa cURL.

Ee, pịa bọtịn egwu na ihuakwụkwọ a ka ịgụnye nlele. I nwere ike ịgụnye ngwe emeredịkachọrọ na ihuakwụkwọ ngwe ka ọsụsọ ma mepụta nlele n'efu na ụda ọbụla.

Chọ̀ọ́ Speaker 1 Ugbu a

Tinye ngwe ọbụla ma gụọ ya site na Speaker 1. N'efu iji.

Kewapụta okwu Akaụntụ