# GPT-4o Audio Transcription

> GPT-4o Audio Transcription is a paid API for AI agents from audio.withzero.xyz, paid per call via MPP, $0.005/call, status down (last checked 2026-10-12, last successful call 2026-06-16).

Transcribes base64-encoded spoken audio (WAV or MP3) to text using OpenAI's GPT-4o Transcribe model

## Facts

- Endpoint: POST https://audio.withzero.xyz/api/v1/transcripts/gpt-4o-transcribe?utm_source=zero.xyz
- Price: $0.005/call
- Payment: MPP
- Status: down
- Last checked: 2026-10-12
- Last successful call: 2026-06-16
- Success rate: 7% of calls made through Zero
- Rating: 1.3 / 5 from 1 review
- Activations on Zero: 16
- Provider: audio.withzero.xyz
- Website: https://audio.withzero.xyz
- Canonical page: https://www.zero.xyz/c/audio-withzero-xyz-gpt-4o-audio-transcription-0ee947c3
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_A9kmcpmENGqycOPdOMQf-

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 audio-withzero-xyz-gpt-4o-audio-transcription-0ee947c3 -d '<json body>'
```

Example prompt: Can you transcribe this MP3 audio file for me? I'll send it as base64 — use GPT-4o Transcribe to turn the speech into text.

## When to prefer this

Choose this endpoint when you need high-accuracy speech-to-text transcription using OpenAI's latest GPT-4o Transcribe model, especially for WAV or MP3 audio up to 15 MB. Prefer this over Whisper-based endpoints when you want GPT-4o's superior language understanding and transcription quality. Ideal for agents working with voice memos, meeting recordings, or spoken commands.

## Known failure modes

- Audio exceeds 15 MB decoded size — returns validation error
- Unsupported audio format provided (not wav or mp3) — returns enum validation error
- Malformed or invalid base64 encoding — returns decoding error
- Audio contains no detectable speech — may return empty transcript
- Network timeout for long audio files

## How this service works

Transcribe spoken audio to text with OpenAI's GPT-4o Transcribe.

## Output

Returns a JSON object containing the full text transcript of the audio, the detected duration of the audio in seconds, the model name used (GPT-4o Transcribe), and the USDC price charged for the request.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "$schema": "https://json-schema.org/draft/2020-12/schema",
 "example": {
  "audio": "UklGRiYAAABXQVZFZm10IBAAAAABAAEAQB8AAIA+AAACABAAZGF0YQIAAAAAAA==",
  "format": "wav"
 },
 "required": [
  "audio",
  "format"
 ],
 "properties": {
  "audio": {
   "type": "string",
   "maxLength": 20971524,
   "minLength": 1,
   "description": "Base64-encoded audio to transcribe (decoded size up to 15 MB)."
  },
  "format": {
   "enum": [
    "wav",
    "mp3"
   ],
   "type": "string",
   "description": "Container format of the audio."
  }
 },
 "additionalProperties": false
}
```

## Response schema (JSON Schema)

```json
{
 "type": "object",
 "$schema": "https://json-schema.org/draft/2020-12/schema",
 "required": [
  "text",
  "seconds",
  "model",
  "price_usdc"
 ],
 "properties": {
  "text": {
   "type": "string",
   "description": "The transcript of the audio."
  },
  "model": {
   "type": "string",
   "description": "The model that produced the transcript."
  },
  "seconds": {
   "type": "number",
   "description": "Detected duration of the input audio in seconds."
  },
  "price_usdc": {
   "type": "string",
   "description": "The amount charged for this request, in USDC."
  }
 },
 "additionalProperties": false
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/audio-withzero-xyz-gpt-4o-audio-transcription-0ee947c3/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from audio.withzero.xyz](https://www.zero.xyz/host/audio.withzero.xyz/llms.txt)