Dọkumenti

Kpọnye Speech to Speech AI n'ime usoroiheomume gị na REST API dị mfe.

Nhazi

Bido na Speech to Speech AI API n'okpuru 5 nkeji.

URL nke okpuru
https://api.speechtospeechai.com/v1
1. Bido ọrụ

Tuleelụ ọbụla bụ POST na kii API gị. Ọrụ id bụ nzaghachi, ọ bụghị ụda.

cURL
curl -X POST https://api.speechtospeechai.com/v1/tts/ \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "text=Hello! Welcome to Speech to Speech AI." \
  -F "model=kokoro" \
  -F "voice_id=af_heart" \
  -F "format=mp3"

{"uuid": "3f2c...", "job_id": "...", "status": "queued"}
2. Nhọrọ maka nsonaazụ

Nwere ike ịjụ maka ọrụ n'oge ọbụla ruo mgbe ọnọdụ ahụ gachara, mgbe ahụ budata result_url.

cURL
curl "https://api.speechtospeechai.com/v1/speech/results/?uuid=3f2c..." \
  -H "Authorization: Bearer YOUR_API_KEY"

{"status": "completed", "result_url": "https://.../output.mp3"}

Nkwenye

Nnọọnye API niile chọrọ nkwenye site na kii API na Authorization isiokwu.

Nkeelu
Authorization: Bearer YOUR_API_KEY
Wepụta kii API gị site na gị Akaụntụ dàshboard.

Nhazi ụda

POST /v1/voice-clone/

Kpọnye ngwe n'ime ụda nke ndesịta ozi ahụ.

Nhazi ahụ (ọtụtụ akụkụ/fomu-data)
Paramita Ụdị Ekwesịrị Ndesịta ozi ndị ahụ
reference_audio file Ee Nnọọ nke ụda ka a klọnye (10-30 sekọnd a na-atụ aro, ruo 60)
text string Ee ngwe a ga-ekwu na ụda ahụ
model string Ónyénwē chatterbox, openvoice, tortoise mọọbụ gpt-sovits
language string Ónyénwē Kóòdù asụsụ (dìfọ́ọ̀ltụ̀: en)
Ndesịta ozi ahụ

Ọrụ id (uuid). Pọlụ GET /v1/speech/results/?uuid= ruo mgbe ọnọdụ ahụ ga-agwụ ma bubata result_url.

Nkọwa na-ekwu

POST /v1/tts/

Gbanye ngwe n'ime ụda okwu.

Nhazi ahụ (ọtụtụ akụkụ/fomu-data)
Paramita Ụdị Ekwesịrị Ndesịta ozi ndị ahụ
text string Ee ngwe a ga-atụgharị (maximum 5000 characters)
model string Ónyénwē kokoro (dìfọ́ọ̀ltụ̀), piper, vits mọọbụ bark
voice_id string Ónyénwē Id vùó site na GET /v1/speech/voices/?model=
language string Ónyénwē Kóòdù asụsụ (dìfọ́ọ̀ltụ̀: en)
speed float Ónyénwē Nhazi nke n'elu ọsọ (0.5-2.0, difọ́ọ̀ltụ̀ 1.0)
format string Ónyénwē wav (dìfọ́ọ̀ltụ̀), mp3, ogg mọọbụ flac. SSML táàbụ̀ a na-ewepụ tupú ịmịnye.
Ndesịta ozi ahụ

Ọrụ id (uuid). Pọlụ GET /v1/speech/results/?uuid= ruo mgbe ọnọdụ ahụ ga-agwụ ma bubata result_url.

Nsọ ka ngwe

POST /v1/transcribe/

Kpọnye ụda n'ime ngwe na oge ojoo.

Nhazi ahụ (ọtụtụ akụkụ/fomu-data)
Paramita Ụdị Ekwesịrị Ndesịta ozi ndị ahụ
file file Ee Faịlụ ụda/video ka a gbanwee
model string Ónyénwē v3-obiobi (dìfọ́ọ̀ltụ̀) mọọbụ turbo
language string Ónyénwē Kóòdù asụsụ, mọọbụ ọbụla iji chọpụta ya (dìfọ́ọ̀ltụ̀)
enable_word_timestamps bool Ónyénwē Tinye oge nke mbụ na oge nke ikpeazụ na okwu ọbụla
callback_url string Ónyénwē URL nke na-eweta nsonaazụ site na POST mgbe ọrụ ahụ gasịrị
Ndesịta ozi ahụ

A ọrụ id (uuid). Pọlụ GET /v1/transcribe/results/?uuid= ruo mgbe ọnọdụ ahụ ga-agwụ; nzaghachi ahụ mgbe ahụ na-eburu ngwe, asụsụ na segments.

Model ndị dịnụ

Nhazi ngwe-ka-asụsụ
Model ID Ndesịta ozi ndị ahụ Ọkachasị maka
kokoro TTS multilingual nke n'elu Ónyénwē, latị́nị̀ ala
piper TTS n'asụsụ ọtụtụ dị nfe (ọnụ ndị ahụ na-echekwala n'ahịa ka ọbụla) Nnukwu renders na 11 asụsụ
vits LJSpeech nakwa 108 VCTK English speakers Aịkọn English
bark Generative TTS, 130 presets na 13 asụsụ Nhazi, n'enyemaka nke n'emegharị (n'okporo ụzọ)
openvoice Ntụgharị ụda na-ahọrọ Ụdịmbanye maka Chatterbox
chatterbox Zero-shot ụda ọbụla na 23 asụsụ; difọ́ọ̀ltụ̀ ụda gbanwere móòdù Kloónịga site na saịmpọn nta
Nsụgharị
Model ID Ndesịta ozi ndị ahụ Ọkachasị maka
whisper-large-v3 OpenAI Whisper Large v3 Nhazi dị elu, asụsụ ndị ọzọ
whisper-turbo Ónyénwē Whisper maka nrụọrụ Ntụgharị n'ụzọ nkịtị

Kóòdù ndehie

_Ọdịdị Nhazi Ndesịta ozi ndị ahụ
invalid_api_key 401 Kii API nke a na-enweghị isi mọọbụ nke enweghị
insufficient_credits 402 Agaghị enwe ike iwepụ ihenhọrọ ndị ahụ n'ihi ọrụ a
rate_limit_exceeded 429 Nhazi ndị ahụ dị n'ebe ahụ
invalid_request 400 Nhazi ahụ nke achọrọ ahụ
model_not_found 404 Model achọrọ abụghị nke ọbụla
voice_not_found 404 Agaghị enwe ụda ahụ achọrọ
file_too_large 413 Faịlụ ahụ a na-ebubata dị n'elu ikikembanye
server_error 500 Ndehie sava