# Speech to Text Transcription (Whisper)

> Speech to Text Transcription (Whisper) is a paid API for AI agents from agentsvc.io, paid per call via x402, $0.01/call, status unknown (last checked 2026-10-17).

Transcribes audio files (voice notes, meetings, podcasts, calls) into text with timestamps using OpenAI Whisper, supporting ~100 languages with automatic language detection

## Facts

- Endpoint: POST https://agentsvc.io/api/v1/proxy/audio-transcribe?utm_source=zero.xyz
- Price: $0.01/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-17
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/speech-to-text-transcription-whisper-24094a07
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_SL-bgsNfDfzIPvfy6UHzJ

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 speech-to-text-transcription-whisper-24094a07 -d '<json body>'
```

Example prompt: Can you transcribe this meeting recording for me? Here's the audio file URL: https://example.com/meeting.mp3 — use the small model for best accuracy and include timestamps so I can find specific moments.

## When to prefer this

Choose this endpoint when you need a fast, cost-effective ($0.01/call) audio-to-text conversion with multilingual support and automatic language detection, and when data privacy matters (audio deleted immediately after processing). Ideal for agents handling voice notes, meeting recordings, or podcasts up to 10 minutes. Prefer the 'base' model for speed and 'small' for accuracy. If your audio exceeds 10 minutes or 24 MB, you'll need to chunk it first.

## Known failure modes

- Audio file exceeds 10-minute or 24 MB limit — request rejected
- Unsupported audio format (not mp3, mp4, m4a, wav, ogg, webm, or flac) — decoding error
- Audio URL is not publicly accessible or returns 404 — fetch failure
- Invalid base64 encoding in audio_base64 — decoding error
- Very noisy or inaudible audio may produce low-quality or empty transcription
- Unsupported ISO 639-1 language code passed explicitly — may default or error
- Payment not provided or insufficient — x402 payment required response

## How this service works

Speech to text for agents: transcribe a voice note, meeting, podcast or call recording into text with timestamps, in about 100 languages with automatic language detection. Runs OpenAI Whisper (small by default, base for speed) on agentsvc.io itself, the audio is deleted after the call. Send audio_url or audio_base64 (mp3, mp4, m4a, wav, ogg, webm, flac), up to 10 minutes and 24 MB per call. Returns text, detected language, duration and segments with start and end seconds.

## Output

Returns a JSON object with the full transcribed text, detected language code (e.g. 'en'), language confidence probability (0-1), total audio duration in seconds, processing time in milliseconds, and an array of segments each containing start time, end time, and text. Also includes the Whisper model used and a license note confirming audio is deleted after processing.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "model": {
   "enum": [
    "small",
    "base"
   ],
   "type": "string",
   "default": "small",
   "description": "small is more accurate, base about twice as fast"
  },
  "language": {
   "type": "string",
   "description": "ISO 639-1 code like en or de, default auto-detect"
  },
  "audio_url": {
   "type": "string",
   "description": "Public URL of the audio file"
  },
  "timestamps": {
   "type": "boolean",
   "default": true,
   "description": "Include segments with start and end seconds"
  },
  "audio_base64": {
   "type": "string",
   "description": "Audio file as base64; or use audio_url"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "data": {
   "text": "This is a sample recording for the agents.io speech to text service. It turns spoken audio into text with time stamps in about 100 languages.",
   "model": "whisper-small",
   "language": "en",
   "segments": [
    {
     "end": 4.4,
     "text": "This is a sample recording for the agents.io speech to text service.",
     "start": 0
    },
    {
     "end": 9,
     "text": "It turns spoken audio into text with time stamps in about 100 languages.",
     "start": 4.4
    }
   ],
   "model_license": "MIT (OpenAI Whisper weights, faster-whisper runtime), runs on agentsvc.io, audio is deleted after the call",
   "processing_ms": 3688,
   "duration_seconds": 9.17,
   "language_probability": 0.998
  },
  "success": true
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/speech-to-text-transcription-whisper-24094a07/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from agentsvc.io](https://www.zero.xyz/host/agentsvc.io/llms.txt)