Tags the selected model understands — click to drop one into your text where it happens:
This model reads plain text, so inline tags are ignored. For tag-based emotion, switch to an expressive model like Orpheus or Bark.
Natural-language delivery cues. Honored by Qwen3-TTS today; other expressive models coming soon.
This model doesn't support style instructions — switch to a model that does (e.g. Qwen3-TTS).
Danish text-to-speech is famously hard because Danish pronunciation has drifted far from its spelling — many consonants soften or disappear, vowels reduce, and whole syllables are swallowed, so the written and spoken forms diverge more than in Swedish or Norwegian. Its signature feature is the stød, a creaky-voice "glottal catch" that is phonemic: it can be the only thing distinguishing pairs like "mor" (mother) from "mord" (murder), so a synthesizer must place it correctly to be intelligible. Danish also has an unusually large vowel inventory — well over a dozen distinct vowel qualities plus the letters æ, ø, and å — and a "soft d" (as in "gade") that sounds nothing like an English or German d. Because so much meaning rides on these subtle features, high-quality Danish synthesis depends on a model trained on native speech rather than rules derived from the orthography.
Sample — Dansk
“Tekst-til-tale-teknologien læser danske sætninger op med en naturlig stemme, så de er nemme at lytte til.”
Native name
Dansk
Speakers
About 6 million speakers; the official language of Denmark and a co-official language in the Faroe Islands and Greenland.
Language family
Indo-European, North Germanic (East Scandinavian)
Script
Latin alphabet with æ, ø, å
Spoken in
Denmark, the Faroe Islands, Greenland, and the Danish minority in northern Germany
Accessibility narration for Danish public-sector and government sites
IVR and phone-system prompts for Danish businesses
E-learning and language-practice voiceover in Dansk
Localized app and e-commerce voice prompts for the Danish market
Danish Text to Speech — FAQ
Danish spelling and pronunciation diverge sharply — consonants weaken or drop and vowels reduce, so the spoken word is often much shorter than the written one. Our voice is trained on native Danish audio to capture these reductions, rather than reading each letter, which would sound stilted and foreign.
Yes. The stød is a phonemic creaky-voice catch that distinguishes words like "mor" (mother) from "mord" (murder) or "hun" from "hund". The model places it where native Danish requires, which is essential for the speech to be understood correctly.
Fully. These three vowels are part of the Danish alphabet (added after z) and carry their own sounds, so keep them in your text rather than substituting ae, oe, or aa. The voice also handles the "soft d" in words like "gade" and "mad".
Yes, which is useful because Danish business and tech writing borrows many English terms. The synthesizer voices embedded English words and product names in place while keeping the surrounding Danish prosody and the characteristic Danish vowel qualities.
