Steam player trend data for AI assistants

Track concurrent player counts and game momentum for any title on Steam. Player trend data reveals which games are growing, which are declining, and when major launches or updates are driving spikes.

Updated 2026-08-19

Trends MCP returns a weekly series of monthly concurrent player interest for a Steam title, scaled 0-100, with about 5 years of history when the game is old enough. The keyword is the store display name, not an App ID. Concurrent players measure live sessions, which is closer to "is this still being played" than units sold. The same MCP connection covers 25+ sources, including app download trends and web traffic data. The free plan includes 100 requests per month. Starter is $19/month.

What monthly concurrent players mean

Steam reports how many people are in a title at once. That number jumps at launch, on major patches, during crossovers, and on weekend evenings. It falls when the content drought hits. Trends MCP keeps a monthly concurrent reading as a weekly history so an assistant can see the launch cliff, the seasonal sale bounce, and the long plateau that live-service games try to hold.

The series is not revenue, ARPU, or review-bomb detection. A free weekend can fill servers without filling the cash register. A dying live-service game can still show a loyal concurrent base that is too small to fund the next season. Analysts treat concurrent as engagement, then bring in other sources for commercial proof.

get_trends draws the history. get_growth compresses 3M and 1Y change for screening a watchlist. Live store charts for mobile sit on get_top_trends (for example App Store Top Free) and do not take a Steam display name. Steam history and mobile charts are complementary, not interchangeable.

Display name as the keyword

Use the name on the Steam store page: Counter-Strike 2, Elden Ring, Balatro. Subtitle noise should match the store card. CS2 may fail if the resolver expects the full display name. App IDs (730) and store.steampowered.com URLs are the wrong keyword type.

Sequels and rebrands are separate titles. Counter-Strike 2 is not Global Offensive. Query the name players see today. If a franchise has several live products, query each display name rather than the franchise word.

Non-Steam stores (Epic, console) are outside this source key. A multi-platform publisher needs Steam for PC concurrent, app downloads for mobile, and web traffic for the marketing site.

Steam Web API versus a trend series

Valve's Web API and store endpoints can return current player counts, app lists, and news for developers who register a key. Building a five-year weekly archive still requires polling, storage, and joining other markets. Rate limits and key policy sit on that path.

Trends MCP does not replace Steamworks telemetry, VAC stats, or workshop APIs. It returns a normalized concurrent series inside the same assistant session that already queries npm, mobile installs, and domains. npm trends is the usual sibling when the question is developer-tool adoption rather than player sessions.

Third-party Steam charts sites already plot concurrent players. Those UIs are human dashboards. Trends MCP is the same family of question in JSON for agents, plus the option to call get_growth across Steam and another source in one turn.

Who uses Steam player curves

Publishers watch retention after the review embargo lifts. Funds that hold publisher equity watch whether a surprise hit is still online six months later. Talent scouts watch genre peers to see if a mechanic is still filling servers. Hardware and peripheral brands watch the biggest live titles as a proxy for where PC spend is concentrating.

Indie teams use the series more modestly: compare a similar title's launch shape before setting a wish-list and discount calendar. The 0-100 index is relative to that game's own peak, so a small roguelike at 100 in its window is not "bigger than" a MOBA at 40.

Live-service calendars leave fingerprints: a season launch is a step up, a content drought is a slide, a surprise crossover is a spike that should fade. If the series never comes back after the spike, the crossover borrowed players rather than adding them. Battle-pass games that hold a plateau through two seasons are doing something a one-week chart cannot show. That is why the 5-year weekly archive is the object to store, not a single current-player ping.

Limits and caveats

Offline single-player games under-report concurrent relative to their true audience because many sessions never look "online." Family sharing, LAN, and cracked copies never appear. Regional peaks can be washed out in a weekly point. Steam Sales create calendar spikes that look like product success and are really price.

Games that leave Steam, change name, or merge app IDs can show a broken series. Treat a cliff as a catalog event until the store page confirms it.

Quota is 100 requests per month free, Starter $19/month. A daily poll of the entire top 1,000 is a dedicated charts product. A weekly growth screen of a 30-title watchlist fits the research quota.

Comparison: game-engagement data

Trends MCP is one MCP connection covering 25+ sources. Steam concurrent sits on the same 0-100 index as mobile downloads and site traffic, with about 5 years of weekly history.

Steam Web API remains the path for live current-player calls, app metadata, and studio-owned telemetry. It is not a multi-source research session.

Specialist Steam analytics vendors add wishlists, revenue estimates, and tag explorer UIs. Teams should check current vendor pricing for those dashboards. Trends MCP supplies the assistant-native concurrent series, not a replacement for a full market-intel workstation.

get_trends

Chart concurrent player history for any Steam game by display name. See the launch spike, post-launch retention curve, and whether a title holds players or bleeds out.

get_trends(keyword='Counter-Strike 2', source='steam', data_mode='weekly')

get_growth

Calculate Steam player growth over 3 months and 1 year. Positive growth well after launch usually means a durable audience, not only a launch spike.

get_growth(keyword='Palworld', source='steam', percent_growth=['3M', '1Y'])

get_ranked_trends

Find which games are growing fastest in concurrent players right now - useful for spotting breakout titles before they dominate headlines or Twitch charts.

get_ranked_trends(source='steam', sort='yoy_pct_change', limit=25)

get_top_trends

See which apps and games are topping App Store and Google Play charts right now - surfaces what consumers are downloading and playing across mobile and PC in real time.

get_top_trends(type='App Store Top Free', limit=20)

Common questions

A weekly series of monthly concurrent player interest for a Steam title, normalized to a 0-100 index, with about 5 years of history when the title is that old. The series reflects how many people are in-game at once, not units sold, wishlists, or review scores.
Pass the game's display name as it appears on the Steam store, for example Counter-Strike 2, Palworld, or Elden Ring. The server resolves that name to the Steam App ID. Numeric App IDs and store URLs are the wrong keyword type for this source.
The number of accounts in-game at the same time, as Steam reports it. Peak concurrent users is a standard live-engagement metric. Trends MCP stores that engagement as a monthly concurrent series and returns it on the same 0-100 index used for other sources.
Studios, publishers, and analysts use the curve to judge post-launch retention, season-pass health, and whether a surprise hit is still filling servers months later. It is a player-attention series, not a revenue model. Pair it with web traffic and mobile downloads when the company ships on more than Steam.
Up to about 5 years of weekly points. Titles newer than that simply have a shorter series. The window is long enough for launch spikes, Steam Sale seasonality, and long-tail retention to be visible.