Text-to-Speech

Overview

Converts text into a synthesized speech audio file using the LemonFox TTS API, with support for custom voice types and output formats. Suitable for use cases such as voice feedback in chat apps, blog narration, and video voiceovers.

Property	Value
Endpoint	POST https://app-dohnbj5by5mp-api-GYX1lzGw01Xa.gateway.appmedo.com/v1/audio/speech
Auth	Bearer Token (platform_managed)
Request Format	application/json
Response Format	Binary audio file (application/octet-stream)
Billing	Billed per request

Request Parameters

Parameter	Type	Required	Description
input	string	Yes	The text content to convert to speech
voice	string	Yes	Voice type, e.g.: heart
response_format	string	No	Output audio format, e.g.: mp3, wav, ogg

Response

The endpoint directly returns binary audio file content, with response header Content-Type: application/octet-stream. There is no JSON wrapper — the response body is the audio file itself.

For complete parameter details and code, see references/speech-api.md.

---

Generation-time Usage (Agent Direct Call)

Use the built-in script for generation-time calls. The script reads INTEGRATIONS_API_KEY from the environment, calls the upstream API, and saves the returned binary audio to the file given by --output.

The Bash tool timeout MUST be set to 600000ms (600 seconds).

python3 <skill-path>/scripts/call_text_to_speech.py \
  --input "Hello, world!" \
  --voice heart \
  --response-format mp3 \
  --output /tmp/output.mp3

On success it prints one JSON line to stdout, e.g. {"status":"succeed","file":"/tmp/output.mp3","bytes":12345}. On failure it prints an error to stderr and exits with a non-zero code. After the file is saved, tell the user the path and how to play it.

For complete parameter details, see the "Generation-time Usage" section in references/speech-api.md.

---

Post-generation Usage (In-app via Edge Function)

In an application, requests are proxied through an Edge Function. The Edge Function stores the binary audio stream to Supabase Storage and returns a persistent public URL to the frontend.

For complete Edge Function code and frontend call patterns, see the "Post-generation Usage" section in references/speech-api.md.