Text to Speech Tools

The Text to Speech provider converts text into spoken audio using ElevenLabs voices and returns base64-encoded MP3 audio. Use it in a workflow whenever you need to generate voice output from text (for example, narrating an answer, producing audio notifications, or building voice responses).

Overview

Operations

Configuration

tts_synthesize

Convert text to speech audio using ElevenLabs voices.

The tool issues a POST request to https://api.elevenlabs.io/v1/text-to-speech/{voiceId} with headers Accept: audio/mpeg, Content-Type: application/json, and xi-api-key: {apiKey}. The request body sends text and model_id.

Outputs

tts_synthesize

• data (json) — Base64-encoded audio with MIME type and format. Object shape: audioBase64 (string, the base64-encoded MP3), mimeType (string, audio/mpeg), format (string, mp3).
• metadata (json) — Synthesis metadata:
  • voiceId (string) — Voice ID used for synthesis.
  • characterCount (number) — Number of characters synthesized.

YAML Example

tts_1:
  type: tts
  name: "Text to Speech"
  inputs:
    operation: "tts_synthesize"
    text: "Hello, welcome to our service!"
    voiceId: "21m00Tcm4TlvDq8ikWAM"
    modelId: "eleven_turbo_v2_5"
    apiKey: "{{TTS_API_KEY}}"
  connections:
    outgoing:
      - target: next-block-id

Tips

• Authentication uses an ElevenLabs API key passed as the xi-api-key header. Store it as an environment variable and reference it with {{TTS_API_KEY}} rather than hardcoding the secret.
• voiceId is required — use a valid ElevenLabs voice identifier such as 21m00Tcm4TlvDq8ikWAM. The value is URL-encoded and trimmed before the API call.
• modelId is optional and defaults to eleven_turbo_v2_5 when omitted; set it only if you need a specific ElevenLabs model.
• The audio is returned as base64-encoded MP3 inside data.audioBase64 (MIME type audio/mpeg). Pass {{tts_1.data.audioBase64}} to a downstream block to store, upload, or play the audio.