# Inference402 Speech-to-Text Transcription

> Inference402 Speech-to-Text Transcription is a paid API for AI agents from inference402.worlddata402.workers.dev, paid per call via x402, $0.02/call, status unknown (last checked 2026-10-20).

Transcribes audio from a public URL into text with timestamps and word count using Whisper Large v3 Turbo, paid per-call in USDC via x402.

## Facts

- Endpoint: GET https://inference402.worlddata402.workers.dev/v1/audio/transcribe?utm_source=zero.xyz
- Price: $0.02/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-20
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/inference402-speech-to-text-transcription-f5508d2e
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_Dpuav2yBR0e0Bf1a7eAhj

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 inference402-speech-to-text-transcription-f5508d2e
```

Example prompt: Transcribe this audio recording for me — it's in Arabic and the file is at https://example.com/interview.mp3 — I need the full text with timestamps.

## When to prefer this

Choose this endpoint when you need keywordless, no-account, pay-per-call audio transcription billed in USDC via the x402 protocol, especially when you want Whisper-quality results with timed segments without managing API keys or subscriptions. Ideal for AI agents that need to process audio on demand without prior account setup.

## Known failure modes

- Audio file exceeds 10MB limit — request rejected
- Audio URL is not publicly accessible or returns 4xx/5xx — fetch failure
- Unsupported audio format — transcription fails
- Invalid ISO 639-1 language code provided — may cause incorrect transcription
- Payment not received or insufficient USDC balance — 402 Payment Required
- Transient Cloudflare Workers AI backend error — service unavailable

## How this service works

Pay-per-call AI inference for agents: chat (Llama 3.3 70B / 3.1 8B), structured JSON, summarize, translate (incl. Egyptian Arabic), extract, classify, sentiment, keywords, rewrite, embeddings, moderation, vision and speech-to-text. USDC on Base mainnet over x402. No API key, no account.

## Output

Returns a JSON object with the full transcribed text, the model used (Whisper Large v3 Turbo), detected or hinted language code, word count, total audio duration in seconds, and an array of timed segments each with start time, end time, and segment text.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "language": {
   "type": "string",
   "description": "Optional ISO 639-1 spoken-language hint, e.g. ar"
  },
  "audio_url": {
   "type": "string",
   "format": "uri",
   "description": "Public http(s) URL of an audio file ≤10MB"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "text": "Hello and welcome ...",
  "model": "@cf/openai/whisper-large-v3-turbo",
  "language": "en",
  "segments": [
   {
    "end": 4.2,
    "text": "Hello and welcome",
    "start": 0
   }
  ],
  "word_count": 96,
  "duration_seconds": 42.1
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/inference402-speech-to-text-transcription-f5508d2e/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from inference402.worlddata402.workers.dev](https://www.zero.xyz/host/inference402.worlddata402.workers.dev/llms.txt)