Pass source google images for visual demand, or source google search for web demand. There is no Google Images live board.
Live data as of 2026-09-20
A google images vs google search comparison on Trends MCP is two keyword sources on get_growth and get_time_series. Source google images returns a 0-100 visual-demand series for any phrase. Source google search returns a 0-100 web-demand series for the same phrase, with an estimated volume field on some keywords. There is no get_top_trends type for Google Images. From the September 20, 2026 Google Trends board (as_of 2026-09-20T06:01:39 UTC), rank 3 was ed sheeran. Dedicated get_growth closed that name at index 66.0 on google search and 1.0 on google images for recent_date 2026-09-19. The image series peaked at 100.0 on 2026-05-02 while search sat at 9. type Google Images is a 400. type google search aliases to Google Trends.
The September 20, 2026 Google Trends board is the split that still matters next month: live rank is a web-search title list, google search can sit at a series peak, and google images can sit near the floor on the same name. Omitting limit returned limit 25, count 25, offset 0, as_of_ts 2026-09-20T06:01:39.935577+00:00. Rank 1 was how do i register to vote. Rank 2 was lizzie borden. Rank 3 was ed sheeran. Rank 15 was national cheeseburger day.
| Token | Trends rank | google search 7D | google images 7D |
|---|---|---|---|
| ed sheeran | 3 | 66.0 from 8.0, +725.0% | 1.0 from 0.0, 999999.0% |
| national cheeseburger day | 15 | 27.0 from 1.0, +2600.0% | 53.0 from 5.0, +960.0% |
| travis kelce | 19 | 9.0 from 3.0, +200.0% | 4.0 from 3.0, +33.33% |
| resident evil movie | 7 | 43.0 from 11.0, +290.91% | 20.0 from 6.0, +233.33% |
| spiderman | off-board | 15.0 from 20.0, -25.0% | 42.0 from 48.0, -12.5% |
| nike | off-board | 56.0 from 59.0, -5.08%, ~20,900,000 estimated | 55.0 from 57.0, -3.51% |
| greenland | 25 | 8.0 from 5.0, +60.0%, ~1,940,000 estimated | 16.0 from 14.0, +14.29% |
Live rank is not a sort of either index. ed sheeran sat at rank 3 with google search 66.0, a 261-week close from 2021-09-25 at 33, and google images 1.0. national cheeseburger day sat at rank 15 with images 53.0 ahead of search 27.0. Both of those holiday rows carried data_quality low and a warning that the underlying signal is over 90% zero values. Quote the indexes. Do not treat 53.0 as a year-round visual brand.
The inverse peak is the durable join. get_time_series on google images for ed sheeran returned 261 weekly rows from 2021-09-25 (index 2) to 2026-09-19 (index 1). Peak 100.0 landed on 2026-05-02. google search on that same week sat at 9. The search series then closed 2026-09-19 at 66, the highest point in those 261 rows, while images sat at 1. 12M on search was 66.0 from 15.0, +340.0%. 12M on images was 1.0 from 1.0, flat. YTD on search was 66.0 from 9.0 on 2026-01-03, +633.33%. YTD on images was 1.0 from 1.0, flat.
spiderman stayed visual after the film window. Images closed 42.0 on 2026-09-19, 30D -38.24% from 68.0, 12M +82.61% from 23.0. Search closed 15.0, 30D -64.29% from 42.0, 12M +87.5% from 8.0. The images series peaked at 100.0 on 2026-08-08, then 80.0, 68.0, 62.0, 55.0, 48.0, 42.0. Search never matched that August 8 height on this pull.
1W matched 7D exactly on ed sheeran for both sources, including the images 999999.0. 1M matched 30D exactly: search 66.0 from 9.0, +633.33%; images 1.0 from 4.0, -75.0%. REST omitting percent_growth on google images defaulted to 12M and printed 1.0 from 1.0, 0.0%, 261 points. A later MCP omit on google search defaulted to 12M and printed 66.0 from 15.0, +340.0%.
Two strings, one tool family. REST uses mode get_growth or get_time_series. MCP uses the matching tool names. Image windows live on the Google Images trends API. Web-search volume lives on Google search volume data. Window defaults live on get-growth. The live web board is documented with google search vs Google Trends.
{"mode": "get_growth", "source": "google images, google search", "keyword": "ed sheeran", "percent_growth": ["7D", "30D", "12M"]}
{"mode": "get_growth", "source": "google search", "keyword": "ed sheeran", "percent_growth": ["7D", "1W", "30D", "1M", "12M", "YTD"]}
{"mode": "get_time_series", "source": "google images", "keyword": "ed sheeran"}
{"mode": "get_time_series", "source": "google search", "keyword": "ed sheeran"}
{"mode": "get_top_trends", "type": "Google Trends"}
The MCP equivalents are get_growth(keyword="ed sheeran", source="google images, google search", percent_growth=["7D", "30D", "12M"]), get_time_series(keyword="ed sheeran", source="google images"), and get_top_trends(type="Google Trends"). REST top_trends with type Google Trends and limit 3 on September 20, 2026 returned the same snapshot timestamp as the omitted-limit MCP pull, with count 3: how do i register to vote, lizzie borden, ed sheeran.
The comma-separated Ed Sheeran call returned sources_requested ["google images", "google search"], sources_successful 2, and trend_analysis.uptrend_ratio 2/2 because both 7D directions printed increase. volume_weighted_growth printed null, weight_coverage 0/2 sources, because neither close carried a usable volume field on that name. Same preset, same recent_date 2026-09-19, different indexes: the earlier multi-source row printed search 67.0, then dedicated get_growth and get_time_series closed search at 66.0. Quote 66.0 from the series last point.
get_time_series on google images, google search returned invalid_source with 16 names and no all. Multi-source strings belong on get_growth only. Source Google Images and source GOOGLE SEARCH both case-folded and returned HTTP 200, echoing google images and google search. Store the lowercase source strings.
how do i register to vote sat at rank 1 with search 100.0 and images 45.0, both from 0.0 on every window including 12M, both tagged data_quality low. lizzie borden sat at rank 2 with search 25.0 from 1.0 and images 15.0 from 1.0, also tagged low. Quote the live rank. Skip a 999999.0 that starts from a zero baseline on a sparse query.
get_top_trends with type Google Images returned HTTP 200 with envelope statusCode 400, error invalid_request, message Unknown type 'Google Images', plus 24 valid feed labels that include Google Trends and do not include any Images board. type google images failed the same way. The handler did not map either string to source google images.
type google search on get_top_trends returned HTTP 200 and echoed type Google Trends, with the same as_of_ts 2026-09-20T06:01:39.935577+00:00 and the same first five titles as the canonical label. type google trends also returned 200 and echoed Google Trends. llms.txt still describes feed labels as case-sensitive. The live handler accepted these lowercase aliases on September 20, 2026. Prefer the canonical Google Trends spelling in stored configs. Do not store google search as a feed type just because it happened to alias.
type google shopping did not get that alias. It returned invalid_request and the same 24-name list. Shopping demand stays on source google shopping for get_growth, the same pattern as images.
get_growth with source Google Trends and keyword ed sheeran returned HTTP 400, error invalid_source, message 'google trends' 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 google search. The get_time_series reject list for a comma-separated pair is the same 16 names without all.
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.
nike is the quiet brand check. Multi-source get_growth closed google search at 56.0 from 59.0 on 7D, -5.08%, volume_estimated true, recent_volume 20,900,000 from 22,000,000. 30D was 56.0 from 73.0, -23.29%, 20,900,000 from 27,200,000. 12M was 56.0 from 54.0, +3.7%, 20,900,000 from 20,100,000. volume_growth was omitted on every window, with reason volume is derived from the trend value for this source, not an independent measurement. google images closed 55.0 from 57.0 on 7D, -3.51%, 55.0 from 61.0 on 30D, 55.0 from 57.0 on 12M, and carried no volume fields. weight_coverage printed 1/2 sources and volume_weighted_growth -5.08, which is the search row only.
greenland sat at board rank 25 with the same volume pattern. Search closed 8.0 from 5.0 on both 7D and 30D, +60.0%, recent_volume 1,940,000 from 1,210,000, volume_estimated true, volume_growth omitted. 12M was 8.0 from 4.0, +100.0%, 1,940,000 from 968,000. Images closed 16.0 from 14.0 on 7D, +14.29%, 16.0 from 21.0 on 30D, -23.81%, 16.0 from 10.0 on 12M, +60.0%, still with no volume field.
ed sheeran search did not print a volume field on the dedicated 6-window call or on the omit-default 12M call. Images never printed one on this pull. Quote 66.0 next to 1.0. Do not invent an image-search count.
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 Google Trends pull on this page is the paid default, which is also what the handler returns when limit is omitted on a paid key. The image follow-up is the Google Images trends API. This split stays inside Google: visual demand versus web demand for one phrase.
FAQ