Pass source google search for a weekly Google Search interest series, or source wikipedia for a monthly page-view series with a volume field. Same string, different grain, different close dates, and a 7D window that only one side can compute.
Live data as of 2026-10-10
Source google search on get_growth and get_time_series returns a weekly 0-100 Google Search interest series for one phrase. Source wikipedia returns a monthly 0-100 page-view series with a volume field for one article title. Both pulls on October 10, 2026 used the same string chatgpt. Google Search closed 72.0 on 2026-10-03 across 260 weekly points from 2021-10-16. Wikipedia closed 34.8 / 4824 on 2026-09-01 across 46 monthly points from 2022-12-01, and echoed the keyword as Chatgpt. The 12M window moved in opposite directions: Google Search minus 20.88 percent from 91.0, Wikipedia plus 13.73 percent from 30.6. 7D computed on Google Search and returned collapsed_range on Wikipedia. The live boards use type Google Trends and type Wikipedia Trending, not those source strings.
The October 10, 2026 pull is the split that still matters next month: the two sources do not share grain or close date, then disagree on 12M direction. get_time_series for chatgpt on source google search returned 260 weekly points from 2021-10-16 (value 0) through 2026-10-03 (value 72). The first non-zero week was 2022-12-10 at value 4. Peak 100 sat on 2025-09-13. No row carried a volume key. The series field is source google search. Growth metadata printed total_data_points 260.
get_time_series for chatgpt on source wikipedia returned 46 monthly points from 2022-12-01 (value 0.0, volume 1) through 2026-09-01 (value 34.8, volume 4824). Peak 100.0 sat on 2025-07-01 at volume 13845. The series field is datatype wikipedia. The keyword came back as Chatgpt. llms.txt still describes about 5 years of weekly history. This source, on this pull, was a 46-month monthly strip that closed a month behind the Google Search week.
Growth windows on that keyword line up by preset name and split by calendar. Google Search 7D matched 1W: 72.0 from 64.0, plus 12.5 percent, recent_date 2026-10-03, baseline 2026-09-26. 14D: 72.0 from 67.0, plus 7.46 percent. 30D matched 1M: 72.0 from 82.0, minus 12.2 percent. 3M was plus 1.41 percent from 71.0. 12M was minus 20.88 percent from 91.0. YTD was plus 12.5 percent from 64.0 on 2026-01-03. Wikipedia 7D, 14D, and 1W each returned collapsed_range with the message that recent and baseline resolved to the same data point. 30D matched 1M: 34.8 from 23.0, plus 51.3 percent, volume 4824 from 3188 (volume_growth plus 51.32 percent), recent_date 2026-09-01. 3M was plus 3.88 percent from 33.5, volume plus 4.15 percent from 4632. 12M was plus 13.73 percent from 30.6, volume plus 13.85 percent from 4237. YTD was plus 0.29 percent from 34.7, volume plus 0.4 percent from 4805. Omitting percent_growth defaults to 12M on both and succeeded on both.
| Keyword | Google close | Google 7D | Google 12M | Wikipedia close | Wikipedia 7D | Wikipedia 12M | Points |
|---|---|---|---|---|---|---|---|
| chatgpt | 72.0 on 2026-10-03 | +12.5% from 64.0 | -20.88% from 91.0 | 34.8 / 4824 on 2026-09-01 | collapsed_range | +13.73% from 30.6 | 260 / 46 |
| tesla | 44.0 on 2026-10-03 | -18.52% from 54.0 | -26.67% from 60.0 | 6.7 / 5340 on 2026-09-01 | collapsed_range | -49.24% from 13.2 | 260 / 60 |
| bitcoin | 24.0 on 2026-10-03 | -33.33% from 36.0 | -46.67% from 45.0 | 5.9 / 100154 on 2026-09-01 | collapsed_range | -45.87% from 10.9 | 260 / 60 |
| Deaths in 2026 | 15.0 on 2026-10-03 | -69.39% from 49.0 | +150.0% from 6.0 | 92.7 / 4107397 on 2026-09-01 | collapsed_range | 999999.0 from 0.0 | 260 / 23 |
| Mathematics | 46.0 on 2026-10-03 | -48.31% from 89.0 | -25.81% from 62.0 | 14.9 / 95476 on 2026-09-01 | collapsed_range | -29.38% from 21.1 | 260 / 60 |
Index and volume are two numbers on Wikipedia. bitcoin 30D index plus 321.43 percent from 1.4. Volume plus 22.19 percent from 81965. Mathematics 30D index printed 999999.0 from 0.0 while volume moved from 78407 to 95476, plus 21.77 percent. Deaths in 2026 12M printed 999999.0 from 0.0 with volume 53 to 4107397. Quote the volume move when the index baseline is 0.0. Do not chart 999999.0 as a percent of page views.
Google Search volume is present on some keywords and missing on others. chatgpt and bitcoin printed none on growth. tesla printed estimated volume 6530000.0 (volume_available true, volume_estimated true, volume_growth null). The series integer on 2026-10-03 was 6530159. Peak 100 / 14841270 sat on 2025-03-15. Wikipedia tesla held 60 monthly points from 2021-10-01 (37.1 / 9014) through 2026-09-01, peak 100.0 / 16616 on 2022-11-01, and printed index 0.0 on 2026-02-01 with volume still 4529. The 0-100 column is not a rescaled view count.
Source names case-fold. Google Search on get_growth for chatgpt returned the same 12M row as the lowercase label. Wikipedia matched the lowercase 12M row. Feed labels do not belong here. Source Google Trends returned HTTP 400 invalid_source, message 'google trends' is not a valid source., and listed 16 keyword sources plus all. Source Wikipedia Trending did the same, echoed as 'wikipedia trending'. The handler lowercased the rejected type string. It did not map it to source google search or source wikipedia.
Google Search chatgpt printed no volume. Wikipedia chatgpt 30D printed real volume on every successful window. Index plus 51.3 percent. Volume plus 51.32 percent. Those are two numbers.
{"mode": "get_growth", "source": "google search, wikipedia", "keyword": "chatgpt", "percent_growth": ["7D", "30D", "12M"]}
That comma call on October 10, 2026 returned sources_successful 2. trend_analysis printed uptrend_count 2, downtrend_count 0, and uptrend_ratio 2/2. Wikipedia 7D had already collapsed. Google Search 7D plus 12.5 percent is an increase. Wikipedia 30D plus 51.3 percent is also an increase. Google Search 30D minus 12.2 percent never entered that 2/2 tally. The mixer skipped the collapsed preset and mixed windows across sources.
volume_weighted_growth was 51.32 with has_volume_data true, sources_with_volume 1, weight_coverage 1/2 sources, and total_volume 4824.0. The only weighted row was Wikipedia, at 100 percent of that volume, using Wikipedia 30D volume_growth 51.32. Google Search 7D plus 12.5 percent never entered. Google Search 12M minus 20.88 percent never entered. Wikipedia index plus 51.3 percent never entered either. The mixer used volume_growth, not the 0-100 index.
A bitcoin 30D comma call printed a cleaner version of the same trap. uptrend_ratio 1/2. volume_weighted_growth was 22.19, again weight_coverage 1/2 sources and total_volume 100154.0. That 22.19 is Wikipedia 30D volume_growth. Wikipedia 30D index plus 321.43 percent stayed out. Google Search 30D minus 35.14 percent stayed out. Read source_results and weight_coverage before charting volume_weighted_growth as a two-platform number.
get_time_series rejects the comma. Source google search, wikipedia returned HTTP 400 invalid_source and listed 16 keyword sources with no all. One source per series call. Two sources belong on get-growth.
Keyword format is the durable split. Source wikipedia wants an article title. On this pull the handler title-cased the input. Keyword chatgpt echoed search_term Chatgpt and matched keyword ChatGPT on the 12M row: 34.8 from 30.6, plus 13.73 percent, volume 4824 from 4237. Keyword ChatGPT (chatbot) returned HTTP 404 no_data. Keyword tesla echoed Tesla and returned the 60-point series above. Keyword Tesla, Inc. returned HTTP 404 no_data. Title-case is not a join key to the company article.
Google Search treats the string as a search phrase. Keyword ChatGPT echoed search_term ChatGPT and matched the lowercase 12M row: 72.0 from 91.0, minus 20.88 percent, 260 points. Store the spelling that was typed if the echo matters. Do not assume Wikipedia's Chatgpt label is a different Google Search series.
A live Wikipedia Trending title is not a guaranteed article series. Rank 1 on type Wikipedia Trending on this pull was Nidal Hasan. That title returned HTTP 404 no_data on source wikipedia. Rank 2 Mike Ditka did the same. Rank 3 Mathematics hit: 14.9 / 95476, 30D volume plus 21.77 percent. Rank 15 Deaths in 2026 hit at 92.7 / 4107397. Google Trends rank 1 hurricane isaias returned HTTP 404 no_data on Wikipedia. Rank 6 nidal hasan on Google Search closed 2.0 with data_quality low and 7D growth 999999.0 from 0.0. Rank is a title list. It is not a page-view sort of last month's series. Title follow-up lives on source wikipedia vs type Wikipedia Trending. Monthly windows live on the Wikipedia page views page.
Two strings, two tools. REST uses mode get_growth or get_top_trends. MCP uses the matching tool names. Google Search history lives on the Google search data page.
{"mode": "get_growth", "source": "google search, wikipedia", "keyword": "chatgpt", "percent_growth": ["30D", "12M"]}
{"mode": "get_time_series", "source": "google search", "keyword": "chatgpt"}
{"mode": "get_time_series", "source": "wikipedia", "keyword": "chatgpt"}
{"mode": "get_top_trends", "type": "Google Trends"}
{"mode": "get_top_trends", "type": "Wikipedia Trending"}
The MCP equivalents are get_growth(keyword="chatgpt", source="google search, wikipedia", percent_growth=["30D", "12M"]), get_time_series(keyword="chatgpt", source="google search"), get_time_series(keyword="chatgpt", source="wikipedia"), get_top_trends(type="Google Trends"), and get_top_trends(type="Wikipedia Trending"). REST 1W matched 7D on Google Search chatgpt. REST 1M matched 30D. The same aliases matched on Wikipedia, where 1W collapsed and 1M matched the 30D row.
Omitting limit on type Google Trends returned limit 25, count 25, offset 0, and as_of_ts 2026-10-10T06:01:41.139355+00:00. Rank 1 was hurricane isaias. Rank 2 was susan dell. Rank 3 was world space week 2026. Those rows are search phrases. They are not article titles. Omitting limit on type Wikipedia Trending returned the same stamp, limit 25, count 25, offset 0. Rank 1 was Nidal Hasan. Rank is live article order, not a sort of the 2026-09-01 Wikipedia index.
REST type google search (the keyword source string, lowercase) returned HTTP 200, echoed type Google Trends, and matched hurricane isaias at rank 1. llms.txt still describes feed labels as case-sensitive. The live handler accepted that lowercase string on October 10, 2026. Prefer the canonical spelling in stored configs.
get_top_trends with type wikipedia returned HTTP 400, error invalid_request, message Unknown type 'wikipedia', plus 24 valid feed labels that include Google Trends and Wikipedia Trending. Source wikipedia is not a feed type. Type Google Trends is not a growth source. Type google search is the trap: it aliases the live board instead of erroring.
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.
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. Keyword follow-up stays on source google search or source wikipedia. The live boards stay on their type strings.
FAQ