MCP ServerTools
get_transcript
Fetch the spoken transcript of a YouTube video.
Returns the language, the video id, and either an array of timestamped segments or a single text blob. Use it when you need the actual words spoken in a video, whether to summarise it, quote from it, or search inside it.
Cost: 1 credit on success.
Parameters
| Name | Type | Required | Default | Description |
|---|---|---|---|---|
video_url | string | yes | - | Video id (dQw4w9WgXcQ) or any URL form (youtube.com/watch?v=…, youtu.be/…, /shorts/…, /embed/…). |
format | "json" | "text" | no | "json" | json returns segments as a list; text returns one concatenated string. |
include_timestamps | boolean | no | true | When format=json, include start + duration (ms). When format=text, prefix each line with [Ns]. |
include_metadata | boolean | no | false | Add a metadata field with title, channel, thumbnail. |
Example call
{
"name": "get_transcript",
"arguments": {
"video_url": "https://youtu.be/dQw4w9WgXcQ",
"format": "text",
"include_metadata": true
}
}Response shape
{
"video_id": "dQw4w9WgXcQ",
"language": "en",
"transcript": "We're no strangers to love ...",
"metadata": {
"title": "Rick Astley - Never Gonna Give You Up",
"author_name": "Rick Astley",
"author_url": "https://www.youtube.com/@RickAstleyYT",
"thumbnail_url": "https://i.ytimg.com/vi/..."
}
}Equivalent REST endpoint: GET /api/v2/transcript.