Create speech, transcribe audio, dub audio, generate sound effects, change voices, or isolate speech through the Zora API. Audio endpoints are asynchronous and require a paid Zora API key or PAYG credit.
PAYG audio usage is charged in VND by completed task usage. Provider credit fields and provider balances are not returned in public Zora responses.
| Model | Use |
|---|---|
whisper-large-v3 | Higher-accuracy transcription and translation |
whisper-large-v3-turbo | Lower-latency transcription |
eleven_v3 | Expressive text-to-speech with inline audio tag control |
eleven_multilingual_v2 | Multilingual text-to-speech |
eleven_flash_v2_5 | Low-latency text-to-speech |
eleven_turbo_v2_5 | Balanced quality and latency text-to-speech |
eleven_multilingual_sts_v2 | Speech-to-speech voice changing |
eleven_text_to_sound_v2 | Prompt-based sound effect generation |
Creation endpoints return a task_id. Poll the task endpoint until status is done, then read Zora output URLs from metadata. Output URLs use /v1/audio/files/{token}and require the same Zora API key to download.
| Field | Required | Description |
|---|---|---|
voice_id | TTS/voice changer | Target voice ID. Use GET /v1/audio/voices to list voices. |
text | TTS/sound effects | Input text or sound effect prompt. |
file | Upload routes | Audio file upload for transcription, dubbing, voice changing, or isolation. |
receive_url | No | Optional webhook URL for task completion. |
target_lang | Dubbing | Target language for dubbed audio. |