Pass source steam for a monthly concurrent-player series, or type Steam Most Played for the live rank board. The two strings are not interchangeable.
Live data as of 2026-09-11
A steam vs Steam Most Played comparison on Trends MCP is two different tools. Source steam on get_growth or get_time_series returns a monthly 0-100 concurrent-player series for one game display name, with a volume field when the pipeline has it. Type Steam Most Played on get_top_trends returns ranked live titles and takes no keyword. From the September 11, 2026 pull (as_of 2026-09-11T04:01:40 UTC), rank 1 was Counter-Strike 2, and that name closed at index 73.8 and volume 1,331,274 on recent_date 2026-08-01. Rank 8 TBH: Task Bar Hero returned HTTP 404 no_data on source steam. Rank 6 FiveM returned source_unavailable. Omit percent_growth and steam growth defaults to 12M, which succeeded on Counter-Strike 2 because the series holds 170 monthly points. Mixing the two strings is a 400.
The September 11, 2026 board is the split that still matters next month: live rank is a title list stamped in September, while source steam closed on 2026-08-01. Omitting limit returned limit 25, count 25, offset 0, as_of_ts 2026-09-11T04:01:40.566161+00:00. Rank 1 was Counter-Strike 2. Rank 2 was Dota 2. Rank 3 was PUBG: BATTLEGROUNDS. Follow-up get_growth calls on source steam used recent_date 2026-08-01, 41 days behind the board stamp:
| Token | Kind | 30D index | 30D volume | 30D change |
|---|---|---|---|---|
| Counter-Strike 2 | rank 1 title | 73.8 from 76.3 | 1,331,274 from 1,420,896 | -3.28% |
| Dota 2 | rank 2 title | 86.0 from 75.1 | 981,123 from 860,019 | +14.51% |
| PUBG: BATTLEGROUNDS | rank 3 title | 19.4 from 19.8 | 771,621 from 792,142 | -2.02% |
| Bongo Cat | rank 4 title | 92.7 from 100.0 | 181,650 from 187,676 | -7.3% |
| Valheim | rank 5 title | 1.6 from 2.6 | 30,973 from 35,085 | -38.46% |
| FiveM | rank 6 title | source_unavailable | source_unavailable | HTTP 502 |
| TBH: Task Bar Hero | rank 8 title | no_data | no_data | HTTP 404 |
| Palworld | rank 9 title | 20.0 from 34.4 | 599,628 from 956,613 | -41.86% |
| Slay the Spire 2 | rank 15 title | no_data | no_data | HTTP 404 |
Live rank is not a sort of the August close. Valheim sat at rank 5 with August volume 30,973. Palworld sat at rank 9 with August volume 599,628. Bongo Cat sat at rank 4 with index 92.7 on 18 monthly points, while PUBG sat at rank 3 with index 19.4 and volume 771,621. Quote both the live rank and the monthly row. Do not treat the 0-100 column as a player-count percent.
FiveM returned HTTP 502 source_unavailable twice, with a SteamCharts 500 in the error body. TBH: Task Bar Hero and Slay the Spire 2 returned HTTP 404 no_data. Those titles still belong on the live table. They cannot support a percent-growth claim on this pull.
sort=rank_change with window=7d and limit 10 used a later stamp, as_of_ts 2026-09-11T06:01:33.845011+00:00, plus baseline_ts 2026-09-04T02:01:41.917420+00:00, then returned count 8. Wallpaper Engine climbed 7 places (18 to 11). PUBG: BATTLEGROUNDS, TBH: Task Bar Hero, and Slay the Spire 2 each climbed 5. window=1d against baseline_ts 2026-09-10T02:01:40.196111+00:00 returned count 6, with Wallpaper Engine at rank 11 from 22. offset 25 with limit 5 returned count 0 and empty data. This board did not continue past row 25 on this pull.
Two strings, two tools. REST uses mode get_growth or get_top_trends. MCP uses the matching tool names. The live board page is the Steam Most Played API. Display-name windows live on Steam trends. Window defaults live on get_growth. Pagination and limit are documented with get_top_trends.
{"mode": "get_top_trends", "type": "Steam Most Played"}
{"mode": "get_growth", "source": "steam", "keyword": "Counter-Strike 2", "percent_growth": ["30D", "3M", "12M", "YTD"]}
{"mode": "get_time_series", "source": "steam", "keyword": "Counter-Strike 2"}
{"mode": "get_growth", "source": "steam, google search", "keyword": "Counter-Strike 2", "percent_growth": ["30D"]}
The MCP equivalents are get_top_trends(type="Steam Most Played"), get_growth(keyword="Counter-Strike 2", source="steam", percent_growth=["30D", "3M", "12M", "YTD"]), and get_time_series(keyword="Counter-Strike 2", source="steam"). REST top_trends with type Steam Most Played and limit 3 on September 11, 2026 returned as_of_ts 2026-09-11T06:01:33.845011+00:00, with count 3: Counter-Strike 2; Dota 2; PUBG: BATTLEGROUNDS. That stamp is later than the omitted-limit MCP pull at 04:01:40. Read as_of_ts before joining two pulls from the same morning.
7D and 1W both collapsed on Counter-Strike 2. 14D and 2W collapsed the same way. 30D matched 1M exactly: index 73.8 versus 76.3, growth -3.28%, volume 1,331,274 from 1,420,896, volume_growth -6.31%. Extra windows still sit on one source-plus-keyword request.
The comma-separated Counter-Strike 2 call returned sources_requested ["steam", "google search"], sources_successful 2, and trend_analysis.uptrend_ratio 0/2. Steam 30D was -3.28% with 170 monthly points and recent_date 2026-08-01. Google Search 30D was -35.14% (37.0 to 24.0) on recent_date 2026-09-05 with 261 weekly points and no volume field in that row. volume_weighted_growth printed -6.31 from steam alone, with weight_coverage 1/2 sources. Same preset, different grain, different close dates. Read source_results before joining those percentages.
get_time_series for Counter-Strike 2 on steam returned 170 monthly rows from 2012-07-01 (index 0.0, volume 1,642) to 2026-08-01 (index 73.8, volume 1,331,274). Each row carried date, value, volume, keyword, and datatype steam. The peak index printed 100.0 on 2023-05-01 with volume 1,802,853. llms.txt describes about 5 years of weekly history. This source, on this pull, was a monthly strip that started in 2012.
Palworld held 32 monthly rows from 2024-01-01 (index 100.0, volume 2,101,535) to 2026-08-01 (index 20.0, volume 599,628). The 12M index jumped 1233.33% from a 1.5 baseline while volume rose 873.94% (61,567 to 599,628). Live rank 9 is a September concurrent list. The 2.1 million launch peak is a January 2024 history fact. Those two numbers do not replace each other.
get_top_trends with type steam returned HTTP 400, error invalid_request, message Unknown type 'steam', plus 24 valid feed labels that include Steam Most Played. type Steam failed with the same error. get_growth with source Steam Most Played and keyword Counter-Strike 2 returned HTTP 400, error invalid_source, message 'steam most played' is not a valid source., plus 16 valid sources: google search, google images, google news, google shopping, youtube, wikipedia, tiktok, reddit, amazon, news sentiment, news volume, app downloads, app rankings, steam, npm, python, all. The handler lowercased the rejected source. It did not map it to steam.
The inverse fold exists on the feed name. type steam most played on get_top_trends returned HTTP 200 and echoed type Steam Most Played, with the same top three titles as the canonical label and as_of_ts 2026-09-11T06:01:33.845011+00:00. llms.txt still describes feed labels as case-sensitive. The live handler accepted this lowercase Steam string on September 11, 2026. Prefer the canonical Steam Most Played spelling in stored configs.
get_time_series on steam, google search returned invalid_source too, with the 16 names and no all. Multi-source strings belong on get_growth only.
REST omitting type is a different 400. POST https://api.trendsmcp.ai/api with mode get_top_trends and limit 5, no type, returned HTTP 200 with envelope statusCode 400, error invalid_request, message The 'type' parameter is required, then the same 24-name list. llms.txt still names that miss missing_parameter. Read statusCode inside the JSON, not only the HTTP line.
country GB with limit 5 on Steam Most Played returned HTTP 400, error invalid_request, message that country is only supported for Google Trends, Google Trends by Category, TikTok Trending Hashtags, TikTok Trending Hashtags by Category, X (Twitter) Trending, and YouTube Trending. Steam Most Played is not on that list. Omit country on this feed.
Trademark suffixes on the board folded when stripped. Apex Legends with the feed's trademark suffix and keyword Apex Legends both closed at index 32.7 and volume 291,356 over 30D, +91.23% from 17.1 / 214,120, on 70 monthly points. Overwatch with the registered-mark suffix and keyword Overwatch both closed at 72.7 / 100,564. HELLDIVERS 2 with and without the feed mark both closed at 12.0 / 144,799, +531.58% from 1.9 / 43,782. EA SPORTS FC 26 with and without the mark both closed at 59.0 / 137,146 from 100.0 / 159,730, with 11 points from 2025-10-01, so 12M returned date_out_of_range. Store the display name. Strip marks if a paste fails.
Omit percent_growth on Counter-Strike 2 and the handler returned HTTP 200 with a single 12M row: index 73.8 versus 83.1, growth -11.19%, volume 1,331,274 from 1,502,245, volume_growth -11.38%, total_data_points 170. The default window is 12M. That default ran. On this title it had enough history. EA SPORTS FC 26 did not: 12M printed date_out_of_range with data_start 2025-10-01 and data_end 2026-08-01. Window length is per title, not a global steam calendar.
YTD on Counter-Strike 2 closed at 73.8 versus 95.8 on 2026-01-01, growth -22.96%, volume 1,331,274 from 1,654,355. 3M closed at 73.8 versus 83.4 on 2026-05-01, growth -11.51%, volume from 1,489,217. Rank 1 on the live board can coexist with a multi-month volume fade. The board does not print that fade.
keyword 730 returned the same 30D math as Counter-Strike 2, and the time series echoed keyword 730 across all 170 rows. keyword CS2 returned the same 30D row. llms.txt still describes the steam keyword as a game display name, with first Steam store search result winning. An App ID can still resolve when store search hits the same title. The live board prints Counter-Strike 2, not 730. Join on the display name.
Other titles on the same recent_date confirm the grain. Marvel Rivals sat at index 6.8 and volume 144,922, down 20.0% over 30D from 8.5 / 196,974, with 21 points; 12M index rose 19.3% while volume fell 3.03%. Wallpaper Engine sat at 88.1 / 117,596, down 2.87% over 30D, and was the largest 7d climber on the rank_change board. Baldur's Gate 3 sat at 10.9 / 83,011, up 12.37% over 30D. Team Fortress 2 sat at 16.4 / 63,267, up 22.39% on the index and down 6.59% on volume. Delta Force, rank 25 on the 04:01 board, sat at 18.6 / 131,021, down 44.14% over 30D from 33.3 / 143,820. Those names are valid keywords. TBH: Task Bar Hero is a valid live rank and still missed the series.
get_top_trends MCP pricing is 1 credit per 10 requested rows, max 10 credits per call. REST counts 1 request per call. Free boards still cap at the top 10. The 25-row pulls on this page are the paid default, which is also what the handler returns when limit is omitted on a paid key. The keyword-window follow-up is the Steam trends page. This split stays inside one platform: a monthly concurrent series versus live player rank.
FAQ