Tag fantatry ny modely voafaritra — tsindrio mba hametrahana iray ao anatin'ny lahatsoratrao izay misy azy:
Mamakiana lahabolana tsotra io modely io, ka tsy raharahaina ny tag anatin'ny andalana. Raha mila fihetseham-po mifototra amin'ny tag ianao, dia miova ho modely maneho fihetseham-po toy ny Orpheus na Bark.
Fanomezana ny teny natoraly. Nanome voninahitra ny Qwen3-TTS androany; modely hafa hanehoana ny tena ho avy tsy ho ela.
Tsy mandray an-tànana ny toro-làlan'ny endriky ny teny ity modely ity — ampio ilay modely izay mandray an-tànana izany (ohatra: Qwen3-TTS).
Mamaritra ny fanononana safidy (teny = fanononana):
-12
+12
Manara-maso ny fanehoana (0 = tsy misy ifandraisany, 1 = tena fanehoana)
Ny lanjany amin'ny famaritana tsy misy mpizara (ny avo kokoa = manaraka ny toro-hevitra kokoa)
Mamaritra ny endriky ny feo amin'ny teny natoraly (ity no ampiasaina amin'ny Parler fa tsy ny feo efa voafaritra mialoha)
Lamin'ny takila Dia: Ampiasao ny tag [S1] sy [S2] mba hamantarana mpiteny samihafa. Ohatra: [S1] Misaotra! [S2] Misaotra, ahoana ny toe-draharaha?
Mandarin text-to-speech lives or dies on tone: it has four lexical tones plus a neutral tone, and getting the contour wrong turns "mā" (mother) into "mǎ" (horse), so the model must predict pitch per syllable, not just per sentence. Tone sandhi adds another layer — for example two third tones in a row shift the first to a rising tone, and the words "一" (yī) and "不" (bù) change tone depending on what follows. Because Hanzi carry no spaces and many characters are polyphonic (多音字), high-quality Chinese synthesis depends heavily on word segmentation and grapheme-to-phoneme disambiguation from context.
Ohatra — 中文（普通话）
“今天天气很好，我们一起去公园散步，顺便买点水果回家吧。”
Anaran'ny fiteny
中文（普通话）
Mpitondra feo
about 1.1 billion speakers (roughly 920 million native Mandarin)
Fiteny
Sinitic branch of Sino-Tibetan
Baiko soratra
Chinese characters (Hanzi) — Simplified and Traditional
Notenenina tao
Mainland China, Taiwan, Singapore, Malaysia, Hong Kong, global Chinese diaspora
Ny zavatra ampiasain'ny olona Chinese (Mandarin) Soratra ho teny miresaka ho an'ny
E-learning and Mandarin language-teaching narration
Short-video (Douyin/Bilibili) and livestream voiceover
Navigation and in-car voice prompts
Customer-service IVR and chatbot voices
News and audiobook narration for the diaspora
Chinese (Mandarin) Soratra ho teny - FAQ
Yes. You can paste either Simplified (mainland/Singapore) or Traditional (Taiwan/Hong Kong) text; both are read in Mandarin pronunciation.
The model predicts each syllable's tone contour from context and applies tone sandhi rules, so sequences like third-tone pairs and the special cases of 一 and 不 come out naturally.
Mostly yes. Multi-reading characters such as 行 (xíng vs háng) or 长 (cháng vs zhǎng) are disambiguated from surrounding words, though rare proper nouns can still be ambiguous.
These voices are Standard Mandarin (Putonghua). Cantonese uses a different tone system and pronunciation and is not the same as Mandarin TTS.
Efa nahatratra ny fetra amin'ny modely maimaimpoana ianao. Ampitomboy ny isa mba hahazoana marika fanampiny, na mividy fonosana marika mba hampiasana modely premium.
