RecordSync Voice

REST API v1

Available on packages with API access. Create a key in Dashboard → API Keys and send it as Authorization: Bearer <key>. Jobs are asynchronous: create one, then poll it until status is completed.

GET https://voice.recordsync.cam/api/v1/me

Account, package and remaining credit.

curl -H "Authorization: Bearer $KEY" https://voice.recordsync.cam/api/v1/me
POST https://voice.recordsync.cam/api/v1/stt

Transcribe audio/video (multipart). Fields: file*, model (small|medium|large-v3-turbo|large-v3), language (bn, en, …, empty = auto), task (transcribe|translate), max_chars, max_lines, initial_prompt.

curl -H "Authorization: Bearer $KEY" -F file=@talk.mp3 -F language=bn -F model=large-v3 https://voice.recordsync.cam/api/v1/stt
GET https://voice.recordsync.cam/api/v1/stt/{id}

Job status. When completed, includes segments and download links.

curl -H "Authorization: Bearer $KEY" https://voice.recordsync.cam/api/v1/stt/{id}
GET https://voice.recordsync.cam/api/v1/stt/{id}/download/{format}

Download srt | vtt | txt | json | docx.

curl -H "Authorization: Bearer $KEY" -o out.srt https://voice.recordsync.cam/api/v1/stt/{id}/download/srt
GET https://voice.recordsync.cam/api/v1/voices

Voices you can use (your cloned voices + stock voices).

curl -H "Authorization: Bearer $KEY" https://voice.recordsync.cam/api/v1/voices
POST https://voice.recordsync.cam/api/v1/tts

Text to speech (JSON). text*, language*, voice (id from /voices), format (mp3|wav), speed 0.5-2, pitch -6..6, exaggeration 0.25-2, cfg_weight 0-1, seed, pronunciations {"word":"say"}.

curl -H "Authorization: Bearer $KEY" -H 'Content-Type: application/json' -d '{"text":"আমার সোনার বাংলা","language":"bn","voice":"VOICE_ID"}' https://voice.recordsync.cam/api/v1/tts
GET https://voice.recordsync.cam/api/v1/tts/{id}

Job status, with audio_url when completed.

curl -H "Authorization: Bearer $KEY" https://voice.recordsync.cam/api/v1/tts/{id}
GET https://voice.recordsync.cam/api/v1/tts/{id}/audio

Download the generated audio.

curl -H "Authorization: Bearer $KEY" -o speech.mp3 https://voice.recordsync.cam/api/v1/tts/{id}/audio
Statuses: queued → processing → completed | failed (credit is refunded) | cancelled. Rate limits: 120 req/min, 20 STT and 30 TTS submissions/min. Errors return JSON with message and HTTP 4xx.