Metinden Söze AI Agentler Musiqa Sözden metin Ses Klonlama Ses Dizaýny Bazar _Ýaşa
Hadysa / Fe'l-aşgabat soragyny habar ber
Darwin TTS

Darwin TTS TTS

A Qwen3-TTS variant whose talker FFN weights are blended from the Qwen3 language model for sharper cross-lingual cloning.

0/500 karakterler · 5000 per generation üçin ýaz →
_Gün tertibi 5000 karakter çäk

Metini SSML taglarda dolap dogry kontrol üçin:

<speak><prosody rate="slow">Slow speech</prosody></speak>

Saýlanan model aňlaýan taglar — birini metinde goýmak üçin basyň:

Bu model ýönekeý metin okaýar, şonuň üçin hatda taglar gözden düşürilýär. Tag-based emotions for, switch to an expression model like Orpheus or Bark.

Öz sözleriň terjimesini belli et (söz = terjime):

-12 +12
0.5x 2.0x
Piper, VITS, MeloTTS bilen azat
Siziň döreden audioňyz şu ýerde görüner. Bir model saýlaň, metin girin we döred
Ses mübärek bejerildi
0:00
Ses ýükle .srt ýükle Baglanyşyk 24 sagadyň içinde gutarýar
Özüni ulan $5/mo-dan söwda lisenziýasy
TTS.ai-ni söýýäňmi? Dostlaryňa aýt!

Habar Darwin TTS

Darwin-TTS-1.7B-Cross by FINAL-Bench is a research variant of Qwen3-TTS-1.7B with an unusual construction: 84 of its talker-FFN tensors (about 8.6% of them) are blended at a 3% ratio with the matching tensors from Qwen3-1.7B-Base, all without any retraining. The result is a model that produces noticeably crisper cross-lingual voice cloning across Korean, English, Japanese, and Chinese — its four core languages. It operates in zero-shot voice-clone mode, needing only about three seconds of reference audio to capture a speaker. Darwin is best suited to transferring a single reference voice across those four languages, for example dubbing or multilingual narration with consistent speaker identity.

Muňa iň gowy: Cross-lingual voice cloning between English / Korean / Japanese / Chinese with a single reference voice

Ehlini _Gözle Darwin TTS sesler

Bir seretseň

Developer
FINAL-Bench
Lisenziýa
Apache 2.0
_Göçür
standard
Tizlik
medium
Ses klonlamak
Eý
Diller
English, Korean, Japanese, Chinese
Maks. karakterler
2000

Darwin TTS sesler

Default

English
_Öň bellenen Neutral

Default (Chinese)

Chinese
_Öň bellenen Neutral

Default (Japanese)

Japanese
_Öň bellenen Neutral

Default (Korean)

Korean
_Öň bellenen Neutral

Darwin TTS TTS - Gynançly Soraglar

Darwin starts from Qwen3-TTS-1.7B but blends a small fraction of its talker-FFN weights with the matching weights from the Qwen3-1.7B base language model. This training-free blend sharpens cross-lingual voice cloning rather than changing the base voices.

English, Korean, Japanese, and Chinese. The FINAL-Bench release specifically markets its cross-lingual blend for those four, and the deployed model ships voices for them.

About three seconds. It works in zero-shot mode, so no fine-tuning or training is required — you provide a short reference clip and it generates new speech in that voice.
← Ehli Sesler