API Documentation

Haɗa Speech to Speech AI zuwa cikin aikace-aikace ɗinku ta amfani da API mai sauƙi na REST.

Um farawa ta gaggawa.

Ka fara amfani da API na Speech to Speech AI cikin minti 5.

URL na farko
https://api.speechtospeechai.com/v1
1. Fara aiki

Duk wani kayan aiki shi POST ne tare da maɓallin API ɗinka. Amsa ita ce ƙididdigar aikin, ba sauti ba.

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. Bincike don gano sakamakon.

Yi tambaya game da aikin kowace sakan kaɗan har sai an kammala halin, sa'an nan a saukar da 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"}

Tabbatarwa.

Duk tambayoyin API suna buƙatar tabbatarwa ta hanyar maɓallin API cikin Authorization header

Shugabanci.
Authorization: Bearer YOUR_API_KEY
Samu lambar ku ta API daga wajen ku. accounts-action.

Voice Cloning

POST /v1/voice-clone/

Karanta wata takarda ta amfani da murya wacce aka adana a cikin wani ƙaramin faifai.

Jiki na tambaya (sashe-da-dabaru/form-data)
Parameters Jinsi. Wajibi. Bayani.
reference_audio file Ee. Reƙo na maganar da za'a kwafa (10-30 sakan an shawarce su, har zuwa 60)
text string Ee. Yi magana da wannan murya.
model string A'a. chatterbox, openvoice, tortoise ko gpt-sovits
language string A'a. Kodi na harshen (tsofafi: en).
Mawallafi

Wani alamar aiki (uuid). Poll GET /v1/speech/results/?uuid= har sai an kammala halin da kuma sauke result_url.

Text to Speech

POST /v1/tts/

Canza rubutun zuwa sauti.

Jiki na tambaya (sashe-da-dabaru/form-data)
Parameters Jinsi. Wajibi. Bayani.
text string Ee. Text to convert (max 5000 characters)
model string A'a. kokoro (farawa), piper, vits ko bark
voice_id string A'a. ID na magana daga GET /v1/speech/voices/?model=
language string A'a. Kodi na harshen (tsofafi: en).
speed float A'a. Mai yawatar gudu (0.5-2.0, default 1.0)
format string A'a. (default), mp3, ogg ko flac. Alamar SSML ana cire su kafin a yi amfani da su.
Mawallafi

Wani alamar aiki (uuid). Poll GET /v1/speech/results/?uuid= har sai an kammala halin da kuma sauke result_url.

Speech to Text

POST /v1/transcribe/

Mayar da sauti zuwa rubutu tare da adadi na lokaci.

Jiki na tambaya (sashe-da-dabaru/form-data)
Parameters Jinsi. Wajibi. Bayani.
file file Ee. Fayil na sauti/ bidiyo da za'a canza
model string A'a. large-v3 (default) ko turbo
language string A'a. Ƙoɗar harshe, ko kuma a gane ta da kai (farawa)
enable_word_timestamps bool A'a. Ƙara lokacin da kalma ya fara da lokacin da ya ƙare a kowane kalma.
callback_url string A'a. URL da za'a karɓa sakamakon ta hanyar POST idan aikin ya ƙare
Mawallafi

Wani alamar aiki (uuid). Poll GET /v1/transcribe/results/?uuid= har sai an kammala halin; amsa ta yi da rubutu, harshe da sassa.

Samfuran da ke nan.

Tsarin canzawa rubutu zuwa magana.
Model ID Bayani. Mafi kyau ga.
kokoro Sauran motsa jiki na motsa labarai da sauri, kuma yana iya yin magana da yawan harsuna. Amfani na gabaɗaya, da ƙarancin jefa.
piper Tsarin motsa labarai (text-to-speech) wanda yake sauƙi kuma yana iya magana da manyan harsuna (amma kawai da muradu da aka siyar da izinin su). Mai sauri renders a cikin harsuna 11
vits LJSpeech da masu magana 108 na harshen Ingilishi daga kungiyar VCTK. Saduwar harshen Ingilishi.
bark TTS mai amfani da fasahar samarwa, da sauran ayyuka 130 a harsunan 13. Farkawar kalamin da ta bambanta da kuma ta nuna ma'ana (ta sauri).
openvoice Ƙara sauti da kuma sauya sauti (idan ya zama dole). Mai Sauya ga Chatterbox
chatterbox Ƙirƙirar murya ta atomatik a cikin harsuna 23; tsarin canzawa murya na asali. Ƙirƙirar kwarya daga ƙaramin ɓangare.
Ƙiraye-ƙiraye don canzawa magana zuwa rubutu.
Model ID Bayani. Mafi kyau ga.
whisper-large-v3 OpenAI Whisper Large v3 Matafiya mafi yawa, da kuma iya magana da manyan yaren duniya.
whisper-turbo An ƙirƙiri tsarin "Whisper" don sauƙaƙe ayyuka da saurin aiki. Saurin rubuta da sauti.

Kodin Kuskura.

Kodi. QDialogButtonBox Bayani.
invalid_api_key 401 Maɓallin API mara kyau ko da ya ɓace
insufficient_credits 402 Ba'a da kuɗaɗen da suka isa ga wannan aiki
rate_limit_exceeded 429 Yawan buƙatun da aka aika.
invalid_request 400 Mai dafataccen bayanin da aka aika.
model_not_found 404 Ba'a da nau'in da aka tambaya
voice_not_found 404 Ba'a da maganar da aka tambaya
file_too_large 413 An kasa fitar da fayil ɗin da aka canza girmansa
server_error 500 Kuskure cikin ma'aikacin