Twitch Videos Scraper — VOD Metadata & Metrics

What is Twitch Videos Scraper — VOD Metadata & Metrics?

Twitch Videos Scraper — VOD Metadata & Metrics returns one dataset row per public Twitch video (VOD) from the channels you list: video ID and link, title, publish time, duration in seconds, view count, broadcast type, game or category, thumbnail, and channel context including display name and follower count. Influencer marketers, esports and gaming analysts, agencies and AI agents use it to compare streamers' recent output, see which games draw views, and build content inventories, with no Twitch account or login.

Why people use this Twitch VOD scraper

• Recent VODs per channel. List channels in channelUrls and get each one's newest archived videos, newest first.
• Views and length together. viewCount and durationSeconds let you compare reach per hour of streaming.
• Category on every video. gameId and gameName show what each stream was about.
• Channel context included. channelFollowerCount, channelDisplayName and channelId ride along on every row, ready for views-to-followers ratios.
• Two caps for cost control. maxVideosPerChannel and maxTotalVideos bound each channel and the whole run.

Best fit for this Actor

• Best fit for creator vetting, sponsorship research, gaming category analysis and VOD inventories across known channels.
• Designed for public archived videos on channels you name, using channel URLs or bare logins such as shroud.
• Use resultPosition to keep each channel's videos in Twitch's order, newest first.
• When you need a channel's profile, live status and top clips, continue with Twitch Channel Scraper.
• For creator outreach, then use Twitch Email Scraper on the same channels to find their public business contacts.

Example workflow: from a creator shortlist to a sponsorship pick

An influencer agency starts with fifteen streamers in channelUrls, with maxVideosPerChannel at 20 and maxTotalVideos at 300. The dataset returns each VOD's viewCount, durationSeconds, gameName and publishedAt, plus each channel's channelFollowerCount. They compute views per hour and views relative to followers, compare how often each creator streams the client's game category, and export a ranked shortlist to the client deck. A weekly schedule on the same list then shows who is trending up.

Input

{
  "channelUrls": ["https://www.twitch.tv/shroud", "pokimane"],
  "maxVideosPerChannel": 20,
  "maxTotalVideos": 100
}

Field	Type	What it controls
channelUrls	array of strings	Public Twitch channel URLs or bare logins. Required.
maxVideosPerChannel	integer	Newest archived videos per channel, from 1 to 100. Default 20.
maxTotalVideos	integer	Cap on videos saved across the run, from 1 to 500. Default 100.

Output

One row per video.

Field	What it contains
videoId, videoUrl, title	Video identity, link and title
description	Video description, when the streamer sets one
publishedAt, durationSeconds	Publish time and length
viewCount	Views at collection time
broadcastType	Twitch video type, such as ARCHIVE
gameId, gameName	Game or category
thumbnailUrl	Thumbnail image
channelId, channelLogin, channelDisplayName, channelUrl, channelFollowerCount	Channel context
input, resultPosition	The channel input and the video's position in its feed
recordType, source, scrapedAt	Record type, data source and collection time

Example row:

{
  "recordType": "twitch-video",
  "source": "twitch-web-graphql",
  "input": "https://www.twitch.tv/shroud",
  "channelId": "37402112",
  "channelLogin": "shroud",
  "channelDisplayName": "shroud",
  "channelUrl": "https://www.twitch.tv/shroud",
  "channelFollowerCount": 11293115,
  "videoId": "2888211641",
  "videoUrl": "https://www.twitch.tv/videos/2888211641",
  "title": "ITS FINALLY GLOBAL LETS PLAY WITH GOOD PING :)",
  "publishedAt": "2026-09-30T17:38:37.000Z",
  "durationSeconds": 22246,
  "viewCount": 386979,
  "broadcastType": "ARCHIVE",
  "gameId": "771448419",
  "gameName": "AION 2",
  "thumbnailUrl": "https://static-cdn.jtvnw.net/cf_vods/d2nvs31859zcd8/d105b5c140067be69936_shroud_319408345559_1790789913//thumb/thumb0-640x360.jpg",
  "resultPosition": 1,
  "scrapedAt": "2026-10-01T01:31:19.111Z"
}

Each run also writes OUTPUT and RUN_SUMMARY records to the default key-value store with the run outcome and totals.

Pricing

This Actor uses Pay per event plus Apify platform usage. Each video saved is one Twitch video scraped event, and each run has one Actor Start event plus the platform usage of the run. For example, five channels with maxVideosPerChannel at 20 are up to 100 video events plus the start event and usage. maxTotalVideos caps the whole run, so it doubles as a spending limit. The live Pricing tab on this page shows the current event prices.

API example

curl -X POST "https://api.apify.com/v2/acts/khadinakbar~twitch-videos-scraper/run-sync-get-dataset-items?token=YOUR_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"channelUrls": ["shroud"], "maxVideosPerChannel": 10, "maxTotalVideos": 10}'

from apify_client import ApifyClient

client = ApifyClient("YOUR_API_TOKEN")
run = client.actor("khadinakbar/twitch-videos-scraper").call(run_input={"channelUrls": ["shroud", "pokimane"], "maxVideosPerChannel": 20})
for video in client.dataset(run["defaultDatasetId"]).iterate_items():
    hours = video["durationSeconds"] / 3600
    print(video["channelLogin"], video["gameName"], round(video["viewCount"] / hours))

Use with AI agents (MCP)

Connect the Actor to Claude, ChatGPT or Cursor through Apify MCP and ask:

Get the last 10 VODs from these three Twitch channels and tell me which games each streamer played and which video had the most views per hour.

The agent passes the channels in channelUrls with maxVideosPerChannel at 10, reads rows with get-dataset-items, groups by channelLogin and gameName, and cites each videoUrl. Each saved video is one event, so maxTotalVideos keeps the cost of the request predictable.

Best results

Starting situation	Helpful action	Expected outcome
A long creator list	Set maxVideosPerChannel low, such as 10	A quick, even sample per channel
Deep history on one channel	Raise maxVideosPerChannel toward 100	More archived videos from that channel
Sponsorship vetting	Divide viewCount by durationSeconds and by channelFollowerCount	Comparable reach metrics
Category research	Group rows by gameName	What each creator streams most
Ongoing tracking	Schedule the same channelUrls weekly	Fresh views and new VODs each run

Builder's note

I built this for the question sponsors ask first: what has this creator streamed lately, and who watched it. I kept channel follower counts on every video row so reach ratios need no second lookup, and the two caps make it easy to sample many creators evenly before going deep on a few.

FAQ

Do I need a Twitch account?

No. The Actor reads public Twitch data with no login.

Which videos are included?

The newest public archived videos on each channel, up to maxVideosPerChannel.

Are view counts live?

They are the counts at collection time. Re-run the same input to refresh them.

Responsible use

This Actor collects public Twitch video metadata. Respect Twitch's terms and applicable privacy laws, and review data carefully before using it to make decisions about individual creators. Twitch is a trademark of its owner; this independent Actor is not affiliated with or endorsed by Twitch.