# fastapi.online Speech-to-Text (STT)

> fastapi.online Speech-to-Text (STT) is a paid API for AI agents from api.fastapi.online, paid per call via x402, $0.005/call, down (last probed 2026-10-12); live alternatives: AgentUtility Video Transcription ($0.1) and medien.halowerk.com Speech-to-Text ($0.02).

Agents call it through Zero, which pays per call: no API key and no wallet of your own. New accounts start with $5.

Live alternatives for the same use case:
- [AgentUtility Video Transcription](https://www.zero.xyz/c/agentutility-video-transcription-whisper-v3-83d051eb) ($0.1)
- [medien.halowerk.com Speech-to-Text](https://www.zero.xyz/c/medien-halowerk-com-speech-to-text-stt-22257e66) ($0.02)

Transcribes audio files (WAV, MP3, OGG, etc.) to text with language detection, timestamps, and segment-level detail, paid per call via crypto.

## Facts

- Endpoint: POST https://api.fastapi.online/stt?utm_source=zero.xyz
- Price: $0.005/call
- Payment: x402
- Status: down
- Last checked: 2026-10-12
- Last successful call: 2026-08-13
- Success rate: 67% of calls made through Zero
- Rating: 5.0 / 5 from 1 review
- Activations on Zero: 3
- Tags: x402
- Canonical page: https://www.zero.xyz/c/fastapi-online-speech-to-text-stt-d9986abb
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_DBddXAR2pkwKbuBfUl4hG

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 fastapi-online-speech-to-text-stt-d9986abb -d '<json body>'
```

Example prompt: Can you transcribe this audio file for me — it's an MP3 of a recorded meeting — and tell me what language it's in and break it down into segments?

## When to prefer this

Choose this endpoint when you need no-signup, pay-per-call speech-to-text with crypto (USDC) and want segment-level timestamps and automatic language detection in a single call. It's ideal for agents operating without a pre-registered account, processing sporadic audio jobs, or needing a self-hosted alternative to cloud STT providers like Whisper API or Google Speech-to-Text.

## Known failure modes

- Unsupported audio format returns an error — stick to WAV, MP3, OGG, or other common formats
- Very long audio files may time out or be rejected if they exceed size limits
- Payment failure or insufficient USDC balance returns a 402 error before processing
- Corrupted or silent audio files may return empty text or fail engine-side
- Low-quality or noisy audio may produce inaccurate transcripts with no error signal

## How this service works

Chat, embeddings, OCR, speech-to-text, image editing, image understanding (captioning/Q&amp;A), text-to-speech, translation, writing tools, classification tools, developer tools, local utilities, format/checksum validators, encoding &amp; crypto utilities, date/time calculators, geo/coordinate math, math &amp; finance calculators, novelty tools, and business/social writing generators — 179 endpoints in all, across multiple specialized self-hosted models. Pay per call with crypto — no signup — or get an API key.

## Output

Returns a JSON object with the full transcribed text string, the detected language code, the audio duration in seconds, the name of the engine used, and an array of timestamped segments for fine-grained alignment.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "file": {
   "type": "string",
   "format": "binary",
   "description": "Audio file (WAV/MP3/OGG/etc)"
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/fastapi-online-speech-to-text-stt-d9986abb/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from api.fastapi.online](https://www.zero.xyz/host/api.fastapi.online/llms.txt)