Sample Outputs¶

This page shows representative outputs for each tubebrain tool and transcript format.

The exact text, timestamps, and metadata depend on the video you query. These examples are intentionally small so you can understand the output shape without having to inspect a full transcript.

get_transcript¶

Example call:

get_transcript(url: "https://youtube.com/watch?v=dQw4w9WgXcQ", format: "json")

JSON¶

{
  "video_id": "dQw4w9WgXcQ",
  "title": "Example Video",
  "channel": "Example Channel",
  "duration_ms": 125000,
  "language": "en",
  "source": "caption_manual",
  "segments": [
    {
      "text": "Hello and welcome.",
      "start_ms": 0,
      "end_ms": 1800
    },
    {
      "text": "This is a short example.",
      "start_ms": 1800,
      "end_ms": 4200
    }
  ]
}

Notes:

• language is the language actually returned, not only the language you requested.
• source tells you which path produced the transcript:
• caption_manual
• caption_auto_generated
• whisper_local

Markdown¶

# Example Video
**Channel**: Example Channel | **Duration**: 2:05 | **Language**: en

## Transcript

**[0:00]** Hello and welcome.

**[0:01]** This is a short example.

SRT¶

1
00:00:00,000 --> 00:00:01,800
Hello and welcome.

2
00:00:01,800 --> 00:00:04,200
This is a short example.

VTT¶

WEBVTT

00:00:00.000 --> 00:00:01.800
Hello and welcome.

00:00:01.800 --> 00:00:04.200
This is a short example.

Plain Text¶

Hello and welcome. This is a short example.

get_transcript_section¶

Use this when the user gives a timestamp or asks about one section of a long video.

Example call:

get_transcript_section(url: "https://www.youtube.com/watch?v=Rzi7oFTzjac&t=2449s", format: "json")

Example output excerpt, with segment text shortened for shape:

{
  "video_id": "Rzi7oFTzjac",
  "title": "We need to talk about gstack",
  "channel": "Nerd Snipe",
  "duration_ms": 4786000,
  "language": "en",
  "source": "caption_auto_generated",
  "anchor_ms": 2449000,
  "window_start_ms": 2329000,
  "window_end_ms": 3049000,
  "segments": [
    {
      "text": "[caption text from the requested section]",
      "start_ms": 2448200,
      "end_ms": 2454300
    },
    {
      "text": "[next caption segment in the requested section]",
      "start_ms": 2454300,
      "end_ms": 2460500
    }
  ]
}

Notes:

• anchor_ms is the timestamp requested by at_s or parsed from URL parameters such as t=2449s.
• window_start_ms and window_end_ms describe the filtered transcript window.
• The segment list keeps original video timestamps so downstream agents can cite or reopen the section precisely.

list_languages¶

Example call:

list_languages(url: "https://youtube.com/watch?v=dQw4w9WgXcQ")

Example output:

{
  "video_id": "dQw4w9WgXcQ",
  "languages": [
    {
      "code": "en",
      "name": "English",
      "is_auto_generated": false,
      "is_translatable": true
    },
    {
      "code": "es",
      "name": "Spanish",
      "is_auto_generated": true,
      "is_translatable": true
    }
  ]
}

Use this tool when language choice matters more than grabbing any available caption track.

get_metadata¶

Example call:

get_metadata(url: "https://youtube.com/watch?v=dQw4w9WgXcQ")

Example output:

{
  "video_id": "dQw4w9WgXcQ",
  "title": "Example Video",
  "channel": "Example Channel",
  "duration_ms": 125000,
  "has_captions": true,
  "caption_languages": [
    "en",
    "es"
  ]
}

Error Output Shape¶

Tool failures are returned as plain strings prefixed with Error:.

Example:

Error: invalid YouTube URL or video ID: not-a-youtube-url

For common failures and what to do next, see Troubleshooting.