YouTube Caption & Whisper AI Transcript Extractor: MCP config, n8n workflow and API

{
  "mode": "videos",
  "videoUrls": [
    "https://www.youtube.com/watch?v=jNQXAC9IVRw"
  ],
  "includeAutoGenerated": true,
  "useWhisper": true,
  "whisperModel": "tiny",
  "includeWordTimings": false,
  "proxyCountry": "GB"
}

---
name: youtube-caption-and-whisper-ai-transcript
description: "One record per video with its transcript (segments, full text, SRT/VTT, optional word timings), caption/translation metadata, chapters, heatmap, and channel identity. Runs the Apify Actor nimbryxo-studio/youtube-caption-and-whisper-ai-transcript. Use when the user needs YouTube data with fields such as video_id, title, channel_name, publishedAt, transcript_language, segment_count and an Apify token is available."
compatibility: "Needs network access, curl, and an APIFY_TOKEN environment variable. Runs are billed to the token owner's Apify account."
metadata:
  actor: "nimbryxo-studio/youtube-caption-and-whisper-ai-transcript"
  store-url: "https://apify.com/nimbryxo-studio/youtube-caption-and-whisper-ai-transcript"
---

# YouTube Caption & Whisper AI Transcript Extractor

Use this skill to fetch YouTube data with the Apify Actor `nimbryxo-studio/youtube-caption-and-whisper-ai-transcript` and work with the returned rows.

## Before running

1. Check that `APIFY_TOKEN` is set in the environment. Never print it or write it to a file.
2. Build `input.json` from the user's request using the fields below. Keep limits small for a first run.
3. Confirm the search terms or URLs with the user if the request is ambiguous.

## Input fields

| Field | Type | Meaning |
| --- | --- | --- |
| `mode` | one of `videos`, `channel`, `playlist` | What to scrape: specific videos, a channel's uploads, or a playlist. |
| `videoUrls` | array | YouTube watch URLs, youtu.be short links, Shorts/embed URLs, or plain 11-character video IDs (mode=videos). Duplicate IDs are processed once. |
| `includeAutoGenerated` | boolean | Whether to include auto-generated captions when manual captions are not available. |
| `useWhisper` | boolean | When YouTube has no transcript (disabled captions, music-only, or blocked), download the audio and transcribe with Whisper AI (faster-whisper, CPU). Adds ~30-180 s per video. |
| `whisperModel` | one of `tiny`, `base`, `small`, `medium`, `large-v3-turbo` | Whisper model size to use when `useWhisper` is enabled. Larger = more accurate but slower and more RAM. 'base' is balanced. |
| `includeWordTimings` | boolean | Add a wordTimings array with word-level start/end times. Only auto-generated YouTube captions expose word timings; manual captions and AI transcripts do not. |
| `proxyCountry` | string | Country/region for the proxy IP. Useful if transcripts are geo-blocked in the default US region. |

Example `input.json` (the Actor's published prefill):

```json
{
  "mode": "videos",
  "videoUrls": [
    "https://www.youtube.com/watch?v=jNQXAC9IVRw"
  ],
  "includeAutoGenerated": true,
  "useWhisper": true,
  "whisperModel": "tiny",
  "includeWordTimings": false,
  "proxyCountry": "GB"
}
```

## Run

```bash
curl -sS -X POST "https://api.apify.com/v2/acts/nimbryxo-studio~youtube-caption-and-whisper-ai-transcript/run-sync-get-dataset-items" \
  -H "Authorization: Bearer $APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d @input.json > results.json
```

The call waits for the run to finish and returns the dataset items as a JSON array. If the run needs more than 300 seconds the API answers with HTTP 408; then start the run with `POST /v2/acts/nimbryxo-studio~youtube-caption-and-whisper-ai-transcript/runs`, poll `GET /v2/actor-runs/{runId}` until `SUCCEEDED`, and read `GET /v2/datasets/{defaultDatasetId}/items`.

## Output

Each item is one record. Published fields include: `video_id`, `title`, `channel_name`, `publishedAt`, `transcript_language`, `segment_count`, `success`, `transcript_available`, `transcript_source`, `accessStatus`, `errorCode`, `duration_seconds`, `scrapedAt`, `canonicalUrl`, `channel_id`, `channelHandle`, `full_text`, `transcript_language_name`, `is_auto_generated`, `was_translated`.

Summarise what was found, cite the record URLs when present, and say how many items came back. Do not invent values for empty fields.