# ToolSnap Audio Transcribe

> ToolSnap Audio Transcribe is a paid API for AI agents from mcp.toolsnap.app, paid per call via x402, $0.02/call, status unknown (last checked 2026-11-20).

Transcribes a public audio file URL into plain text or SRT subtitle format using speech-to-text

## Facts

- Endpoint: POST https://mcp.toolsnap.app/x402/tools/audio_transcribe?utm_source=zero.xyz
- Price: $0.02/call
- Payment: x402
- Status: unknown
- Last checked: 2026-11-20
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/toolsnap-audio-transcribe-f9ef8b6a
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_1gWfUT9I4W4S4vdync7KY

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 toolsnap-audio-transcribe-f9ef8b6a -d '<json body>'
```

Example prompt: Can you transcribe this podcast MP3 for me — https://example.com/episode42.mp3 — it's about 45 minutes long (2700 seconds), in English, and I'd like the output as plain text.

## When to prefer this

Use this endpoint when you need to convert a publicly accessible audio file into text or SRT captions and want a pay-per-use, serverless approach without setting up a speech-to-text account. It is ideal for agents handling podcast transcription, meeting notes, subtitle generation, or making audio content searchable. Prefer this over self-hosted or account-based STT services when you need immediate access via a single MCP connection with per-call micropayment pricing.

## Known failure modes

- Audio URL is not publicly accessible or returns 4xx/5xx — transcription fails
- Unsupported audio format (e.g. ogg, flac) — rejected by input validation
- Incorrect duration_seconds estimate causing mispricing or insufficient payment
- Language code not supported — may fall back to English or return error
- Very long audio files may time out or require batching
- Payment insufficient if duration estimate is too low

## How this service works

47 server-side tools for AI agents through a single MCP connection. 36 free, no new accounts. Median 98.1% token savings on web content. Paid tools from $0.02.

## Output

Returns the transcribed content as either plain text (the full spoken content as a readable paragraph or paragraphs) or SRT format (numbered subtitle blocks with start/end timestamps and corresponding text). The response is the raw transcript string in the requested format.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "format": {
   "enum": [
    "text",
    "srt"
   ],
   "type": "string",
   "default": "text",
   "description": "\"text\" (default) or \"srt\" (subtitle format with timestamps)."
  },
  "language": {
   "type": "string",
   "default": "en",
   "description": "Language code (e.g. \"en\", \"es\"). Default \"en\"."
  },
  "audio_url": {
   "type": "string",
   "description": "Public audio URL (mp3, mp4, mpeg, m4a, wav, or webm)."
  },
  "duration_seconds": {
   "type": "number",
   "description": "Your best estimate of the audio's duration in seconds — required to quote/charge this call."
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/toolsnap-audio-transcribe-f9ef8b6a/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from mcp.toolsnap.app](https://www.zero.xyz/host/mcp.toolsnap.app/llms.txt)
