# Venice AI Video Transcription

> Venice AI Video Transcription is a paid API for AI agents from api.venice.ai, paid per call via x402, $10/call, status down (last checked 2026-09-15).

Transcribes the audio from a YouTube video URL into text, returning the transcript and detected language.

## Facts

- Endpoint: POST https://api.venice.ai/api/v1/video/transcriptions
- Price: $10/call
- Payment: x402
- Status: down
- Last checked: 2026-09-15
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/venice-ai-video-transcription-af89bc1e
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_sx5sC7vQsk57pl_5LyI9v

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 venice-ai-video-transcription-af89bc1e -d '<json body>'
```

Example prompt: Can you transcribe this YouTube video for me and give me the full text of what's said? Here's the URL: https://www.youtube.com/watch?v=dQw4w9WgXcQ — return it as plain text please.

## When to prefer this

Use this endpoint when you need to convert YouTube video audio into text and want a simple URL-in, transcript-out API with no local infrastructure. Prefer this over general speech-to-text APIs when the source is a YouTube URL rather than an uploaded audio file.

## Known failure modes

- Invalid or inaccessible YouTube URL returns an error
- Video has no audio or only background music — transcript may be empty or inaccurate
- Private or age-restricted YouTube videos may fail to process
- Unsupported video format or non-YouTube URL may return an error
- Insufficient balance returns 402 Payment Required with top-up instructions
- Very long videos may exceed processing limits

## How this service works

Video transcription.

## Output

Returns a JSON object containing the full transcribed text from the video's audio and a detected language code (e.g. 'en'). If text format is requested, returns the transcript as a plain string.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "required": [
  "url"
 ],
 "properties": {
  "url": {
   "type": "string",
   "example": "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
   "description": "YouTube video URL to transcribe."
  },
  "response_format": {
   "enum": [
    "json",
    "text"
   ],
   "type": "string",
   "default": "json",
   "example": "json",
   "description": "The format of the transcript output, in one of these options: json, text."
  }
 },
 "description": "Request to transcribe a YouTube video URL to text.",
 "additionalProperties": false
}
```

## Response schema (JSON Schema)

```json
{
 "type": "object",
 "required": [
  "transcript"
 ],
 "properties": {
  "lang": {
   "type": "string",
   "example": "en",
   "description": "Detected language code for the transcript."
  },
  "transcript": {
   "type": "string",
   "description": "The transcribed text from the video."
  }
 },
 "description": "Video transcription response."
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/venice-ai-video-transcription-af89bc1e/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from api.venice.ai](https://www.zero.xyz/host/api.venice.ai/llms.txt)