deAPI icon

deAPI

Generate images, videos, transcribe audio and boost prompts using deAPI AI models

Actions12

Audio → Generate a Speech

AI-generated

Summary

Converts text to speech audio using various AI models (Kokoro, Chatterbox, Qwen3) with configurable language, voice, and output format.

Inputs

  • model (required) — The AI model to use for speech synthesis (Kokoro, Chatterbox, Qwen3 TTS Custom Voice, or Qwen3 TTS Voice Design).
  • text (required) — The text to generate speech for.
  • kokoroLang (required) — Language for speech synthesis (en-us, en-gb, fr-fr, hi, it, pt-br, es). Only visible when model is Kokoro.
  • chatterboxLang (required) — Language for speech synthesis (ar, zh, da, nl, en, fi, fr, de, el, he, hi, it, ja, ko, ms, no, pl, pt, ru, es, sw, sv, tr). Only visible when model is Chatterbox.
  • qwen3Lang (required) — Language for speech synthesis (Chinese, English, French, German, Italian, Japanese, Korean, Portuguese, Russian, Spanish). Only visible for Qwen3 models.
  • kokoroVoiceEnUs (required) — Voice to use for Kokoro en-us (e.g., af_alloy, am_adam).
  • kokoroVoiceEnGb (required) — Voice to use for Kokoro en-gb (e.g., bf_alice, bm_daniel).
  • kokoroVoiceEs (required) — Voice to use for Kokoro es (e.g., em_alex, ef_dora).
  • kokoroVoiceFrFr (required) — Voice to use for Kokoro fr-fr (e.g., ff_siwis).
  • kokoroVoiceHi (required) — Voice to use for Kokoro hi (e.g., hf_alpha, hm_omega).
  • kokoroVoiceIt (required) — Voice to use for Kokoro it (e.g., im_nicola, if_sara).
  • kokoroVoicePtBr (required) — Voice to use for Kokoro pt-br (e.g., pm_alex, pf_dora).
  • qwen3VoiceChinese (required) — Voice to use for Qwen3 TTS Custom Voice in Chinese (Aiden, Dylan, Eric, Ono_Anna, Serena, Sohee, Uncle_Fu, Vivian).
  • qwen3Voice (required) — Voice to use for Qwen3 TTS Custom Voice in other supported languages (Aiden, Dylan, Eric, Ono_Anna, Ryan, Serena, Sohee, Uncle_Fu, Vivian).
  • instruct (required) — Natural language description of desired voice characteristics (e.g., 'A warm, friendly female voice'). Only visible when model is Qwen3 TTS Voice Design.
  • options.speed — Speech rate multiplier (0.5 to 2.0). Only visible when model is Kokoro.
  • options.format — Audio output format (mp3, flac, wav). Default is mp3.
  • options.waitTimeout — Maximum time to wait for completion in seconds (30-600). Default is 120.

Output shape

No body returned — success/failure only.

The operation returns the original input item with pairedItem metadata; audio is generated asynchronously via a webhook callback. The node waits for completion based on waitTimeout.

Examples

Example 1: Generate English speech using the Kokoro model with a default voice.

Model: Kokoro, Text: 'Hello, welcome to our service.', Language: English (US), Voice: Alloy (Male)

Example 2: Generate speech in Spanish using the Kokoro model with a specific voice.

Model: Kokoro, Text: 'Buenos días.', Language: Spain (es), Voice: Alex (Male)

Example 3: Generate speech in Japanese using the Qwen3 Custom Voice model.

Model: Qwen3 TTS Custom Voice, Text: 'こんにちは。', Language: Japanese, Voice: Vivian

Links

Discussion