Valorant & CS2 Match Data API – Round-by-Round Esports Timeline avatar

Valorant & CS2 Match Data API – Round-by-Round Esports Timeline

Pricing

from $180.00 / 1,000 video processed (per started 15 minutes)s

Go to Apify Store
Valorant & CS2 Match Data API – Round-by-Round Esports Timeline

Valorant & CS2 Match Data API – Round-by-Round Esports Timeline

Get Valorant and CS2 match data from broadcast VOD links (YouTube, Twitch). Returns one row per round: map, round number, score after the round, winner and VOD timestamp, plus map and match rows; optional VLR.gg check. $1.00 per VOD hour ($0.25 per 15 min); unreadable VODs are free. · by TubeText

Pricing

from $180.00 / 1,000 video processed (per started 15 minutes)s

Rating

0.0

(0)

Developer

TubeText Labs

TubeText Labs

Maintained by Community

Actor stats

0

Bookmarked

2

Total users

1

Monthly active users

3 hours ago

Last modified

Categories

Share

TubeText · Esports & sports data · Pay per result. No login. Ready for AI agents. All tools

Quick start (no coding)

  1. Click Try for free at the top of this page. A free Apify account is enough – it includes monthly credit.
  2. In Broadcast VOD links, paste one YouTube or Twitch VOD link of a Valorant or CS2 match per line.
  3. Click Start. About 10 minutes per VOD hour. To test quickly, set Start at / End at to one map.
  4. When it says Succeeded, open the Output tab and click Export → Excel (or CSV). One row per round (map, round, score after the round, winner, VOD timestamp), plus one row per map and one per match.

Cost: $0.25 per started 15 minutes of video processed ($1.00 per VOD hour) – one 45-minute map ≈ $0.75. VODs that can't be opened are free.

Common mistakes

  • CS2 winners show as "A"/"B" → add the two team names in Team names.
  • A round is missing → it ended while the broadcast showed a replay; it's counted in gapBefore, never guessed.
  • A live stream link → not supported yet; use the VOD after the match.

Questions or problems? Open an issue (we read every one). Need a custom version? Ask in an issue or email us – we build it as a private Actor on Apify.

Use from Claude, Cursor or any MCP client

Add https://mcp.apify.com?tools=tubetext/esports-match-timeline as a remote MCP server and sign in to Apify (Claude: Settings → Connectors → Add custom connector). Claude Code:

$claude mcp add --transport http tubetext "https://mcp.apify.com?tools=tubetext/esports-match-timeline"

Then ask, for example: "Get the round-by-round score of every map in this Valorant VOD: [link]" or "List the VOD timestamps where each map of this CS2 final starts, with the final scores: [link]"

🆓 Free to try: Apify's free plan includes $5 of usage every month, no credit card needed – enough for about 5 VOD hours with this tool.

What you get – real output from a test run, shortened. One row per round (plus map and match rows) – here map 2 of VIT vs 100T, VALORANT Champions Lower Final:

MapRoundScoreWinnerVOD time
Summit1VIT 1 – 0 100TVIT0:47:12
Summit4VIT 2 – 2 100T100T0:52:00
Summit11VIT 6 – 5 100TVIT1:07:27

Valorant and CS2 match data API

Call it from your back end with the Apify API. One request runs the Actor and returns the rows as JSON:

curl -X POST "https://api.apify.com/v2/acts/tubetext~esports-match-timeline/run-sync-get-dataset-items?token=$APIFY_TOKEN" \
-H 'Content-Type: application/json' -d '{"videoUrls":["https://www.youtube.com/watch?v=MZjaOtjtRlU"],"startSeconds":2690,"endSeconds":3590}'
  • Big batches: Processing takes about 10 minutes per VOD hour and the sync endpoint waits up to 5 minutes, so sync calls suit short windows (one map or less, via startSeconds/endSeconds). For full VODs, start the run with POST /v2/acts/tubetext~esports-match-timeline/runs and a webhook, then fetch the dataset.
  • Monitoring: Tournament coverage: schedule the run after each match day with the new VOD links; every round row has a VOD timestamp to link back to. Add an Apify webhook to get notified when each run finishes.
  • Cost at volume: 20 VOD hours a month ≈ $20 at the free-plan price ($0.25 per started 15 minutes). VODs that can't be opened are free.
  • Stable output: field names don't change between runs. Failed or empty items come back as status rows that aren't billed, and every row has charged: true/false.
  • Why not the official API: Official game APIs don't cover pro broadcasts round by round for every event. This reads the broadcast scoreboard itself, so it works on VODs of events without a public data feed.

What does Esports Match Timeline do?

Paste the link of a Valorant or CS2 broadcast VOD on YouTube or Twitch and get the match as data: one row per round with the map, round number, score after the round, the round winner and the second in the VOD where it happened, plus a summary row per map and per match. It reads the scoreboard that the broadcast shows on screen, so it isn't limited to events with an official data feed: we tested it on global, regional and Twitch productions (see Accuracy). No login, no game API key.

Who it's for

  • Analysts and coaches: jump straight to round 14 of map 2 in a long VOD, or build a round-by-round dataset of an opponent's recent series for review.
  • Researchers: momentum, comeback and pacing studies across many matches, with every round tied to a VOD timestamp you can check.
  • Media and content teams: timestamped chapters and highlight lists for long broadcasts, and score graphs for recaps.
  • Developers and AI agents: one call returns structured match progress (see the API section).

What data can you get?

  • 🎯 Round rows: map, mapName, round, scoreA, scoreB, winner, firstSeenAt (seconds) and firstSeenAtHms ("0:52:00"), gapBefore (rounds the broadcast didn't show before this one), inferred (true when a round was worked out rather than seen) and evidenceFrames.
  • 🗺️ Map rows: final score, winner, start and end time in the VOD, roundsSeen vs roundsTotal, gapRounds, finalInferred, mapComplete, and validation when you give a VLR.gg link.
  • 🏆 Match row: maps played, maps won per team, frames read and hudCoverage (the share of sampled frames where the score HUD was visible).
  • ✅ Optional check against VLR.gg (Valorant): paste the match page and each map's rounds are compared with VLR's round history.

Accuracy (measured)

Tested on 4 full productions (October 2026): VCT Champions (YouTube), Game Changers Americas (YouTube), VCT EMEA (Twitch) and an ESL CS2 final (YouTube).

ProductionMaps found / realRounds correctWrong scores
Game Changers Americas, Valorant, YouTube5 / 595 / 960
ESL Pro League final, CS2, YouTube3 / 366 / 690
VCT Champions, Valorant, YouTube (one map)1 / 123 / 230
VCT EMEA final, Valorant, Twitch4 / 482 / 820
Total13 / 13266 / 271 (98.2%)0 in about 6,500 frames

The 5 missed rounds ended while the broadcast cut away (replays, interviews); they're flagged with gapBefore, never guessed. On both Valorant matches with a VLR.gg check, every map matched VLR's round history.

Input

  • Broadcast VOD links (videoUrls): YouTube videos or Twitch VODs, one per line.
  • Game (game): auto, Valorant or CS2.
  • Team names (teams): recommended for CS2, whose broadcast HUD often shows logos instead of names.
  • Start at / End at (startSeconds, endSeconds): process only part of a VOD (and pay only for that part).
  • Seconds between frames (sampleSeconds): default 5.
  • VLR.gg match link (validateWith): optional check for Valorant.
  • Include raw per-frame reads (includeFrameRows): for debugging; off by default.

Output

Real rows from VIT vs 100T, VALORANT Champions Lower Final, map 2 (Summit), shortened:

{"rowType": "round", "game": "valorant", "teamA": "VIT", "teamB": "100T", "map": 2, "mapName": "Summit",
"round": 4, "scoreA": 2, "scoreB": 2, "winner": "100T", "firstSeenAt": 3120.62, "firstSeenAtHms": "0:52:00",
"gapBefore": 0, "inferred": false, "evidenceFrames": 2}
{"rowType": "map", "map": 2, "mapName": "Summit", "finalA": 13, "finalB": 10, "winner": "VIT",
"finalInferred": true, "mapComplete": true, "startAtHms": "0:44:57", "endAtHms": "1:34:31",
"roundsSeen": 22, "roundsTotal": 23, "gapRounds": 0}

Here the broadcast never showed the final 13-10 on the scoreboard, so the last round is inferred: true and the map row says finalInferred: true.

How much does it cost?

$0.25 per started 15 minutes of video processed ($1.00 per VOD hour, $0.25 minimum) at the Apify free-plan price, cheaper on higher plans. A 2.5-hour best-of-three costs about $2.50; limit it to one map with Start at / End at and you pay for that part only. If a run hits its time limit or is aborted, you get the rounds read so far and pay only for the 15-minute blocks that were read. Free: VOD links that can't be opened (private, deleted, live, Twitch subscriber-only), VODs our datacenter IPs can't fetch (try again later), runs stopped by your cost limit, and any failure on our side (AI outage, internal error). Each case gets an explanation row with costCase: free. Processing fee ($0.02 per started 15 minutes read, at cost): only when the AI read the video but no scoreboard or round was found (for example a non-match video, or a window inside a single round). It is never added on top of a normal charge. Processing takes about 10 minutes per VOD hour.

Limits

  • Team scores and rounds only: no economy, kills, players or agents in v1.
  • Gaps when the broadcast cuts away: a round that starts and ends during a replay or an interview is counted in gapBefore but not shown.
  • The final round of a map may be inferred when the broadcast skips the final scoreboard. It's always flagged.
  • CS2: pass teams, because many CS2 HUDs show logos rather than names.
  • Side changes without team names: on a CS2 HUD with logos only, a 6-6 half can leave the side assignment ambiguous; the map row then says sideAmbiguous: true. Passing teams doesn't change this, but names let you read the rows. (Valorant overtime without readable tags isn't covered either; Valorant HUDs show tags, so this hasn't come up.)
  • Not supported yet: live streams, Twitch subscriber-only VODs, one match split across several VOD links.

Ready-made examples

TubeText Labs tools

TubeText Labs builds 33 pay-per-result Apify Actors for video, social and e-commerce data – no login, JSON/CSV/Excel output, and ready for ChatGPT and Claude (Apify MCP), n8n, Make and Zapier. All tools: apify.com/tubetext

Transcripts

Video & ad analysis

  • Video Breakdown AI – hook analysis, shot list, cut pacing, products shown and script of any TikTok, Reel or YouTube video
  • Shot-by-Shot Video Analysis Pro – director-level storyboard of any video: framing, camera move, emotion and why each shot works, plus replay peaks, danmaku and comments tied to shots
  • YouTube & TikTok Comments API – comments with like-weighted sentiment and themes, plus the moments viewers react to on long YouTube and Bilibili videos
  • Video Ad Analyzer – competitors' YouTube video ads by brand or domain (Google Ads Transparency), torn down by AI
  • Meta Ad Library Scraper – Facebook and Instagram ads by brand or keyword: video URLs, images, ad copy, CTA and landing page

Social, apps & trends

  • YouTube Channel Analytics – subscribers, recent video performance, Social Blade grade, ranks and estimated earnings
  • Bilibili Scraper – Bilibili (B站) keyword search, popular board, top-100 ranking, video stats, danmaku text and creator (UP主) followers – no login
  • China Trending Topics – Douyin, Weibo, Baidu, Toutiao and Bilibili hot-search boards in one run, in English – plus an optional AI briefing on why topics trend
  • WeChat Article Scraper – WeChat 公众号 articles by keyword or link: full text, English titles and optional English summaries
  • Google Play Reviews + AI Insights – Android app reviews by star rating, country and language, plus an AI report of complaints and praise
  • Steam Reviews Scraper + AI Insights – Steam game reviews with hours played and votes, review score and current players, plus an AI report of top complaints, praise and feature requests
  • Social Blade Scraper – Social Blade stats for YouTube, TikTok, Instagram and Twitch creators – grade, ranks, 2-week growth, daily history, earnings estimates and top-100 lists
  • Spotify Play Count Scraper – play counts for every track of an album, artist or playlist (incl. Top 50 charts), plus monthly listeners, followers, world rank and top cities

E-commerce & travel

  • TikTok Shop Product Scraper – US TikTok Shop prices, items sold, stock per SKU and shop stats
  • AliExpress Search Scraper – AliExpress search results by keyword: price, items sold, rating, shipping, badges
  • Google Hotels Scraper – hotel prices for any destination and dates, per booking site and as a 90-day price calendar
  • Taobao & Tmall Product Scraper – Taobao/Tmall products by Chinese or English keyword, category or URL: CNY prices, shops and seller scores
  • Alibaba & Yiwu Supplier Finder by Image – product photo in, similar Alibaba.com and Yiwu wholesale listings out: price, MOQ, supplier, country and rating
  • Amazon Product Search Scraper – Amazon search results, Best Sellers lists and product pages: price, rating, reviews, monthly sales signal, badges and specs
  • Google Flights Scraper – flight prices for any route or multi-city trip, a cheapest-day calendar, price insights and booking options per site
  • Agoda & Booking.com Hotel Reviews + AI – Agoda hotel reviews plus verified Booking.com reviews shown on Agoda, hotel scores and sub-scores, and an AI report of complaints, praise and action items
  • Gas Prices Scraper (US & Canada) – live gas station prices by ZIP, city or coordinates – regular, premium, diesel, E85 – with an area summary of the cheapest stations
  • Idealo Price Scraper – every shop offer for a product on idealo – price, shipping, delivery, shop rating – in Germany, Austria, France, Italy, Spain and the UK, with a summary per market

Esports & sports data

  • Esports Match Timeline (this tool) – Valorant and CS2 round-by-round scores, winners and VOD timestamps read from broadcast VODs

Questions or problems? Open an issue (we read every one). Need a custom version? Ask in an issue or email us – we build it as a private Actor on Apify.

Found it useful? A short review on this page helps others find it.

Built by TubeText Labs · Pay per result. No login. Ready for AI agents.

FAQ

Which tournaments work? Broadcasts with a visible score HUD. We tested VCT (global and EMEA), Game Changers and ESL Pro League; other productions usually work too, and hudCoverage tells you how much of the VOD had a readable scoreboard. Does it use the game's API or an account? No. It reads the public broadcast video, the same thing a viewer sees. Can I check the results? Every round has a VOD timestamp, so you can jump to it. For Valorant, add the VLR.gg match link and each map is compared automatically. Is it legal? The Actor reads publicly available broadcast VODs without logging in and outputs match facts (scores and timestamps), not video. You're responsible for using the data in line with the platforms' terms and your local law; for unusual uses, ask a lawyer. How long does it take? About 10 minutes per hour of VOD.

Support

Found a bug, or need another game or field? Open an issue – we read every one. Need a custom version? Ask in an issue or email us – we build it as a private Actor on Apify.