# SingIt Speech - Text-to-Speech via Multiple Voice Models

> SingIt Speech - Text-to-Speech via Multiple Voice Models is a paid API for AI agents from ask.singitai.app, paid per call via x402, $1.301/call, status unknown (last checked 2026-10-30).

Converts text to spoken audio using one of 11+ voice models (Kokoro, ElevenLabs, Gemini, xAI, MiniMax, etc.), billed per request via x402 micropayments in USDC

## Facts

- Endpoint: POST https://ask.singitai.app/v1/audio/speech?utm_source=zero.xyz
- Price: $1.301/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-30
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/singit-speech-text-to-speech-via-multiple-voice-models-d44f4389
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_L1MBQrF4lybNI4w7-9ohR

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability singit-speech-text-to-speech-via-multiple-voice-models-d44f4389 -d '<json body>'
```

Example prompt: Read this aloud using ElevenLabs: 'Welcome to our store. Today's specials are 20% off all electronics.' — give me back an MP3.

## When to prefer this

Choose this endpoint when you need text-to-speech without managing API keys or subscriptions across multiple TTS providers. Ideal for agents that need pay-per-use voice synthesis billed via x402 micropayments in USDC, access to a broad selection of voice models in one OpenAI-compatible interface, and no upfront account setup.

## Known failure modes

- Invalid or unsupported voice model name returns an error
- Insufficient USDC balance causes payment failure and no audio is returned
- Empty or missing input text returns a validation error
- Unsupported audio format request may default to MP3 or return an error
- Network or provider-side failure from the underlying TTS provider (ElevenLabs, Gemini, etc.)

## How this service works

SingIt Speech: Text read aloud by 11 voice models (Kokoro, ElevenLabs, Gemini, xAI, MiniMax and more; list at GET /v1/audio/models), no API key needed. Call it when an agent needs spoken audio made from text. OpenAI speech request format, mp3 by default.

## Output

Returns an MP3 audio file (or other requested format) of the input text spoken by the chosen voice model. No audio is generated or charged if there is a processing failure. The response follows OpenAI's speech endpoint format.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "input": {
   "type": "string",
   "maxLength": 4096,
   "minLength": 1,
   "description": "The text to read aloud."
  },
  "model": {
   "type": "string",
   "examples": [
    "venice-kokoro-tts"
   ],
   "description": "A speech model id from GET /v1/audio/models."
  },
  "speed": {
   "type": "number",
   "maximum": 4,
   "minimum": 0.25,
   "description": "Speaking speed."
  },
  "voice": {
   "type": "string",
   "examples": [
    "af_heart"
   ],
   "maxLength": 60,
   "description": "A voice of the model; they are listed at GET /v1/audio/models."
  },
  "response_format": {
   "enum": [
    "mp3",
    "opus",
    "aac",
    "flac",
    "wav",
    "pcm"
   ],
   "type": "string",
   "default": "mp3",
   "description": "The audio format."
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/singit-speech-text-to-speech-via-multiple-voice-models-d44f4389/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from ask.singitai.app](https://www.zero.xyz/host/ask.singitai.app/llms.txt)