StyleTTS 2
Free.ai (self-hosted)
·
tts
·
~500 tokens per clip
·
4.6
kubva 5 Vashandisi vemutauro wechiGerman munyaya iyi
StyleTTS 2 is a Text-to-speech voice built by Yinghao Aaron Li. Strongest at Voice cloning without training data; expressive multilingual speech.. Self-inotarisirwa pa Free.ai GPUs - inofamba zvakasununguka pazuva rako re token pool (500 tokens pa clip). Yakaburitswa pasi peMIT - kushandiswa kwekutengesa kwakabvumirwa pasi peFree.ai.
Semuenzaniso mamodheru
Zvimwe zvinobvunzwa kakawanda
StyleTTS 2 inotsigira akawanda matauro. Iyo yakazara bhokisi rinoenderana nechirongwa; fomu pane iyi peji rinogamuchira chero tenzi uye chirongwa chinopa mazita ezvinyorwa mumitauro yacho. Ona /voice/ kune yakazara multi-engine picker kana iwe uchida chirongwa chekutaura chine chirongwa chekutaura.
Zvizhinji zvema engines zvinopa American English sezvavanoita zvakajairika, uye nenzvimbo-inowirirana accent yezvimwe rurimi. Premium engines zvinogona kukonzera accent variants - nyora mifananidzo yekuenzanisa.
SSML rutsigiro runogona kuchinja zvichienderana nechigadzirwa. Kuramba, prosody, uye emphasis tags zvinopihwa pazvigadzirwa zve premium uye pazvigadzirwa zve self-hosted. Plain text always works - no markup required.
Streaming TTS inowanikwa pa premium engines kuburikidza ne /v1/tts/ API endpoint ne stream=true. Web UI pane iyi peji inodzosera iyo yakazara clip mushure mekudzokorora kupera.
StyleTTS 2 inoita basa paGPU yedu. Generation inotora kubva pazuva nezuva remahara pool yekutanga. Kana yasara, yakabhadharwa tokens kutanga pa $ 5 → 200,000 tokens. Roughly ~ 5 tokens per character, minimum 100 per clip.
Kusvika 5,000 mavara pamubvunzo pawebhu UI. Kuti uwane zvinyorwa zvirefu (mabhuku emashoko, mabhuku akazara), shandisa /voice/audiobook/ izvo zvinosvibisa uye zvinosvibisa otomatiki, kana kuti udza API kuti iite.
Yeah - POST a list of strings to /v1/tts/batch/, or use the workspace UI at /workspace/ to chain TTS into a longer pipeline (e.g., translate → speak → stitch).
Yes — POST text to /v1/tts/ with model="StyleTTS 2" (or the slug on this page). Returns WAV or MP3. See /api/ for full reference + SDK snippets.
Iyi peji inoshandisa kushandura mazwi kuita mashoko, kwete kushandura mashoko kuita mashoko - mashoko anoshandiswa sezvakajairika. Kuti uite shanduro dzemashoko (kuisa paInternet mashoko aunoda kushandura), ona /voice/clone/, iyo inoita kuti iwe uwane kodzero dzemashoko kana kuti uwane mvumo yakanyorwa.
Self-hosted injini kutenderera pa Free.ai-inotungamirwa GPUs; hapana chinosiya yedu servers. Premium injini kutumira mameseji ku upstream model vanopa pasi yedu DPA. Isu hatina kudzidzisa pa yako inputs uye havatenge data.
Yeah - Free.ai inopa kushandiswa kwezvigadzirwa zvemavhidhiyo zvekutengesa. Iyo injini yepasi pechigadzirwa chemushandisi (Apache 2.0, MIT, kana mamiriro ekutengesa) inoratidzwa pamusoro pechigadzirwa uye pane peji remufananidzo; mukuita izvi zvinoreva kuti voiceovers, mavhidhiyo,
Yeah - kukundikana mabasa auto-refund kune chikonzero (zuva nezuva pool kana akabhadhara tokens). Kana mari yekudzosera haasi kuoneka zuva rimwe chete, email contact@free.ai.