ElevenLabs (Kie.ai) icon

ElevenLabs (Kie.ai)

Generate speech, sound effects, and dialogue using ElevenLabs via Kie.ai API

Actions6

Text-to-Speech

AI-generated

Summary

Convert input text into speech audio using specified voice, model, and optional voice characteristics, returning an audio generation task result.

Inputs

  • Model — Choice of text-to-speech model: 'Turbo 2.5' or 'Multilingual v2'. Defaults to 'Turbo 2.5'.
  • Text (required) — The text string to convert to speech.
  • Voice — Select the voice persona for the speech output from a predefined list (e.g., Rachel, Aria, Roger). Defaults to 'Rachel'.
  • Stability — A number from 0 to 1 governing the voice stability for the TTS output. Default is 0.5.
  • Similarity Boost — A number from 0 to 1 that adjusts the voice similarity to the target speaker. Default is 0.75.
  • Style Exaggeration — A number from 0 to 1 controlling style exaggeration level; higher values may increase latency. Default is 0.
  • Speaker Boost — Boolean to boost clarity and target speaker similarity. Defaults to false.
  • Language Code — Language code for the TTS output (e.g., 'en' for English). Defaults to auto-detection if left blank.
  • Reply URL — Optional webhook URL to be called when the speech synthesis task completes.
  • Reply Ref — Custom string returned in the webhook callback to identify the request.
  • Wait for Completion — Boolean indicating whether to wait for the speech synthesis task to complete before returning (polls every 3 seconds, 5 minutes timeout). Defaults to true.

Output shape

a list of JSON objects containing the task creation response or final task result when wait is enabled

If 'Wait for Completion' is true, the operation polls the task status and returns the final result when the speech generation finishes. If false, it immediately returns the task creation response containing the task ID. The output includes URLs to the generated audio in the response data.

Links

Discussion