Tagi wybrany model rozumie — kliknij aby usunąć jeden do swojego tekstu, gdzie się to dzieje:
Model ten czyta tekst zwykły, więc w linii tagi są ignorowane. Dla emocji na tag, przełącz na model wyrażony jak Orfeus lub Bark.
Wskazówki dostarczania naturalnego języka. Zaszczycony przez Qwen3-TTS dzisiaj; wkrótce nadchodzą inne modele ekspresyjne.
Model ten nie obsługuje instrukcji stylowych — przełącz na model, który ma (np. Qwen3-TTS).
Definiuj własny wymówki (słowo = wymówka):
-12
+12
Kontroluje ekspresywność (0 = neutralna, 1 = bardzo wyrażona)
Waga kierownicza bez klasyfikatora (wyższa = więcej szybkości następującej)
Opisz styl głosu w języku naturalnym (Parler używa tego zamiast ustawionych głosów)
Format dialogów Dia: Użyj [S1] i [S2] tagi do oznaczenia różnych głośników. Przykład: [S1] Witaj! [S2] Cześć, jak się masz?
Portuguese text-to-speech has to commit to a variety, because Brazilian (pt-BR) and European (pt-PT) Portuguese diverge sharply in pronunciation. Brazilian is more vowel-forward and palatalizes "ti/di" toward "chi/dji," while European Portuguese reduces unstressed vowels heavily and is often perceived as more clipped. Both share heavy nasalization (ão, õe, -m endings) and the marked phonemes that the engine must render from diacritics like ã, ç, and the circumflex. The bulk of TTS volume targets the much larger Brazilian market — fintech apps, e-learning, and IVR systems — while pt-PT serves Portuguese and African Lusophone audiences with their own broadcast standard.
Próbka — Português
“Olá, bom dia! Esta é uma demonstração da síntese de voz em português, que transforma texto escrito em fala natural.”
Co ludzie używają Portuguese tekst do przemówienia
Brazilian-Portuguese IVR and chatbot voices for fintechs and PIX flows
E-learning and EAD course narration for Brazilian online universities
European-Portuguese audio for Portuguese broadcasters and government services
Audiolivros and accessibility audio for Lusophone publishers in Brazil and Portugal
Navigation and delivery-app voice prompts for iFood, Rappi, and ride-hailing services
Portuguese Tekst do wypowiedzi — FAQ
Yes. Brazilian (pt-BR) and European (pt-PT) voices differ in vowel reduction and the "ti/di" sounds, so pick the variant matching your audience. The same written text is read with the pronunciation rules of the chosen variety.
Yes. Nasal diphthongs (pão, mãe, corações) and nasal endings (-m, -ns) are core to Portuguese, and the engine renders the nasalization natively rather than as plain vowels.
All diacritics — á, â, ã, à, ç, é, ê, í, ó, ô, õ, ú — are read correctly. They mark stress, vowel quality, and nasality, so keeping them in your text is essential for accurate pronunciation.
The synthesizer reads whatever spelling you provide. Post-reform forms (e.g. dropping the c in "ato") and older spellings are both pronounced correctly, since the change is mostly orthographic rather than phonetic.
