# ADH Agent-native Speech-to-Text (STT)

> ADH Agent-native Speech-to-Text (STT) is a paid API for AI agents from agentdatum.com, paid per call via x402, $0.01/call, status down (last checked 2026-10-13).

Transcribes base64-encoded audio (MP3/WAV/M4A) into text with language detection, returning the transcript and identified language code.

## Facts

- Endpoint: POST https://agentdatum.com/api/v1/tool/tool-stt?utm_source=zero.xyz
- Price: $0.01/call
- Payment: x402
- Status: down
- Last checked: 2026-10-13
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/adh-agent-native-speech-to-text-stt-e33941f8
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_YvDm2-eCvzVpNKo3J85rb

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 adh-agent-native-speech-to-text-stt-e33941f8 -d '<json body>'
```

Example prompt: Can you transcribe this base64-encoded MP3 audio clip for me and tell me what language is being spoken? Here's the audio data: [base64 string].

## When to prefer this

Choose this endpoint when you need a pay-per-call ($0.01 USDC via x402) speech-to-text transcription that works with base64-encoded MP3, WAV, or M4A audio and also returns the detected language code — especially suitable for AI agents operating within the ADH agentdatum.com ecosystem that already use x402 payment settlement.

## Known failure modes

- Invalid or malformed base64 audio data returns an error
- Unsupported audio format beyond mp3/wav/m4a may fail
- Very long audio files may exceed processing limits
- Low-quality or noisy audio may produce inaccurate transcriptions
- Missing required audio_b64 field returns a validation error

## How this service works

ADH 金融情报层：把 239 个原始数据源洗成带实体解析、时间同步、溯源与派生结论的情报（宏观regime / 加密尽调 / 事件窗口），免费注入你的 AI 上下文。Agent 走 x402 结算，人类订阅（PayPal）即将上线。

## Output

Returns a JSON object with 'text' (the full transcription of the audio) and 'language' (a language code such as 'zh' or 'en' indicating the detected language of the audio).

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "text": "transcribed text",
  "language": "zh"
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/adh-agent-native-speech-to-text-stt-e33941f8/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from agentdatum.com](https://www.zero.xyz/host/agentdatum.com/llms.txt)