get-youtube-video-with-transcript.js

import { Innertube } from "youtubei.js";

/**
 * Fetches details and transcript for a YouTube video.
 * @param {string} videoId - The YouTube Video ID.
 * @return {Object} Video details and transcript.
 */
async function getVideo(videoId) {
  const youtube = await Innertube.create();
  const info = await youtube.getInfo(videoId);

  let transcript = null;

  try {
    const transcriptData = await info.getTranscript();
    const segments = transcriptData.transcript.content.body.initial_segments;

    if (segments && segments.length > 0) {
      transcript = segments
        .filter((segment) => segment.snippet && segment.snippet.text)
        .map((segment) => ({
          text: segment.snippet.text,
          start_ms: segment.start_ms,
        }));
    }
  } catch (transcriptError) {
    console.warn(
      "Could not retrieve transcript. The video may not have captions enabled."
    );
    console.warn("Error message:", transcriptError.message);
  }

  return {
    id: info.basic_info.id,
    title: info.basic_info.title,
    author: info.basic_info.author,
    channel_id: info.basic_info.channel?.id ?? null,
    duration: info.basic_info.duration,
    view_count: info.basic_info.view_count,
    transcript: transcript,
  };
}

// Example Usage:
try {
  const video = await getVideo("jNQXAC9IVRw");
  console.log(video);
} catch (error) {
  console.error("An error occurred while fetching the video:", error);
}

// Example Response:
// {
//   id: 'jNQXAC9IVRw',
//   title: 'Me at the zoo',
//   author: 'jawed',
//   channel_id: 'UC4QobU6STFB0P71PMvOGN5A',
//   duration: 19,
//   view_count: 375769099,
//   transcript: [
//     {
//       text: 'All right, so here we are, in front of the\nelephants',
//       start_ms: '1200'
//     },
//     {
//       text: 'the cool thing about these guys is that they\nhave really...',
//       start_ms: '5318'
//     },
//     { text: 'really really long trunks', start_ms: '7974' },
//     { text: "and that's cool", start_ms: '12616' },
//     { text: '(baaaaaaaaaaahhh!!)', start_ms: '14421' },
//     {
//       text: "and that's pretty much all there is to\nsay",
//       start_ms: '16881'
//     }
//   ]
// }