# FetchHarbor Audio Transcription

> FetchHarbor Audio Transcription is a paid API for AI agents from fetchharbor.benlab.download, paid per call via x402, $0.025/call, status unknown (last checked 2026-09-14).

Transcribes up to five minutes of base64-encoded audio locally, returning detected language, duration, and transcript text without retaining any data.

## Facts

- Endpoint: POST https://fetchharbor.benlab.download/audio/transcribe
- Price: $0.025/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-14
- Activations on Zero: 1
- Tags: x402
- Canonical page: https://www.zero.xyz/c/fetchharbor-audio-transcription-4543743a
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_kurKUdvDX0k0ka4qyw74C

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 fetchharbor-audio-transcription-4543743a -d '<json body>'
```

Example prompt: Can you transcribe this base64-encoded audio clip for me — it's under five minutes, spoken in English — and give me the full transcript text along with the detected language and duration?

## When to prefer this

Choose this endpoint when you need privacy-preserving, local transcription of short audio clips (up to five minutes) encoded as base64, especially when data retention is a concern. It is ideal for agents processing sensitive voice recordings since neither audio nor transcripts are stored. Prefer it over cloud ASR services when privacy, cost predictability, or offline-style processing is important.

## Known failure modes

- Audio exceeds the operator-defined size limit — returns an error indicating decoded size is too large
- Invalid or malformed base64 encoding — returns a decoding error
- Unsupported audio format or corrupted audio data — returns a processing failure
- Audio contains no intelligible speech — may return an empty transcript
- Language hint is unrecognized — may fall back to auto-detection or return a validation error

## How this service works

Transcribe up to five minutes of base64 audio in the isolated local media worker. Returns detected language, duration, and text without retaining the source audio or transcript.

## Output

Returns the detected spoken language code, the duration of the audio in seconds, and the full plain-text transcript of the speech content. No audio or transcript is retained after processing.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "language": {
   "type": [
    "string",
    "null"
   ],
   "maxLength": 16
  },
  "audio_base64": {
   "type": "string",
   "description": "Base64-encoded audio; decoded size is operator limited.",
   "contentEncoding": "base64"
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/fetchharbor-audio-transcription-4543743a/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from fetchharbor.benlab.download](https://www.zero.xyz/host/fetchharbor.benlab.download/llms.txt)