Export to JSON and CSV

Every model (Video, Comment, VideoDetails, Transcript, …) can serialize itself. Search and comment paginators expose the same helpers — no list() needed. Module-level dumps_json / dumps_csv still work on any iterable.

from ytscrape import YouTube

with YouTube() as yt:
    results = yt.search("python", max_results=10)
    results.dump_csv("search.csv")
    print(results.to_json())

    details = yt.video("dQw4w9WgXcQ")
    details.dump_json("video.json")

    thread = yt.comments("dQw4w9WgXcQ", max_results=50)
    thread.dump_csv("comments.csv")
    print(thread.to_json())

Async paginators use the same names, awaited:

async with AsyncYouTube() as yt:
    results = await yt.search("python", max_results=10)
    await results.dump_csv("search.csv")
    print(await results.to_json())

CLI:

ytscrape search "python" --max 5 --format json
ytscrape comments dQw4w9WgXcQ --max 20 --format csv -o comments.csv
ytscrape video dQw4w9WgXcQ --format json -o video.json

JSON includes computed fields such as url. Search items also get a type key (video / channel / playlist). Transcript CSV writes one row per caption snippet.

Subtitle files

Fetched transcripts also expose to_srt() / to_vtt() and dump_srt(path) / dump_vtt(path) for SubRip and WebVTT subtitles:

with YouTube() as yt:
    transcript = yt.transcript("dQw4w9WgXcQ", languages=["en"])
    transcript.dump_srt("captions.srt")
    transcript.dump_vtt("captions.vtt")

These helpers accept file paths or text streams. See the transcript guide for timing and empty-caption handling.