Pass source tiktok for a hashtag volume series, or type TikTok Trending Hashtags for the live board. The two strings are not interchangeable.
Live data as of 2026-09-08
A TikTok search vs TikTok Trending Hashtags comparison on Trends MCP is two different tools. Source tiktok on get_growth or get_time_series returns a 0-100 hashtag-volume series, with absolute volume when the pipeline has it. Type TikTok Trending Hashtags on get_top_trends returns ranked hashtag labels and takes no keyword. From the September 8, 2026 pull (as_of 2026-09-08T06:01:33 UTC), rank 1 was tiktokshop, and that token returned HTTP 404 no_data on source tiktok. Rank 17 paidpartnership printed index 100.0 and volume 229,422,842. Omit percent_growth and tiktok growth defaults to 12M, which then fails because the series is 30 daily points. Mixing the two strings is a 400.
The September 8, 2026 omitted-country board is the split that still matters next month: live rank is not a volume ranking, and many board labels have no keyword series at all. Omitting limit returned limit 25, count 25, offset 0, as_of_ts 2026-09-08T06:01:33.729790+00:00. Rank 1 was tiktokshop. Rank 2 was movie. Follow-up get_growth calls on source tiktok used recent_date 2026-09-05 and 30 daily points:
| Board rank | Hashtag | 7D index | 7D volume | 7D change |
|---|---|---|---|---|
| 1 | tiktokshop | no_data | no_data | HTTP 404 |
| 6 | tiktokshopbacktoschool | 49.3 from 88.8 | 136,605,150 from 201,805,961 | -44.48% |
| 7 | onthisday | 84.7 from 22.9 | 111,917,197 from 100,510,405 | +269.87% |
| 10 | nfl | 57.8 from 85.3 | 154,413,688 from 172,536,780 | -32.24% |
| 13 | sohochic | 100.0 from 70.7 | 105,560,501 from 82,115,789 | +41.44% |
| 15 | targetaudience | 94.0 from 71.4 | 119,433,813 from 113,066,927 | +31.65% |
| 17 | paidpartnership | 100.0 from 4.1 | 229,422,842 from 76,906,592 | +2339.02% |
| 22 | nyc | 74.2 from 13.2 | 97,688,063 from 83,595,602 | +462.12% |
| 24 | lindsayclancy | 41.7 from 14.2 | 134,326,266 from 95,127,220 | +193.66% |
| 25 | dollyparton | 8.3 from 33.5 | 55,191,695 from 219,861,882 | -75.22% |
Thirteen of the 25 board labels returned no_data on the first source tiktok try, including rank 1 tiktokshop, rank 2 movie, rank 4 football, rank 5 ad, rank 8 school, rank 9 comedy, rank 11 tiktokgrowthchallenge, rank 12 liveincentiveprogram, rank 14 halloween, rank 16 college, rank 18 makeup, rank 19 cute, and rank 21 minecraft. Rank 3 dance printed index 100.0 and volume 242,407,094 on the first 7D call, then a later retry in the same run returned no_data. Treat a miss as a miss. Do not invent a volume for the rank 1 token.
Among the labels that stayed up, paidpartnership at rank 17 held about 1.5 times the 7D volume of nfl at rank 10, and still sat seven places lower. dollyparton at rank 25 had fallen 75.22% in a week on 55,191,695 volume. onthisday at rank 7 had risen 269.87% on 111,917,197. The board is a ranked label list. The series is a short daily index with a separate volume column.
sort=rank_change with window=7d and limit=10 still used the same as_of_ts, plus baseline_ts 2026-09-01T02:01:32.812429+00:00. The first mover was collegefootball at rank 54 (prev_rank 136, rank_change 82). Other movers sat at ranks 87 through 162. Mover ranks routinely sit past the default 25-row window.
Two strings, two tools. REST uses mode get_growth or get_top_trends. MCP uses the matching tool names. The keyword series page is TikTok trends. The live board page is the TikTok Trending Hashtags API. Window defaults live on get_growth. Pagination and limit are documented with get_top_trends.
{"mode": "get_top_trends", "type": "TikTok Trending Hashtags"}
{"mode": "get_growth", "source": "tiktok", "keyword": "nfl", "percent_growth": ["7D", "1W", "14D", "30D", "1M"]}
{"mode": "get_time_series", "source": "tiktok", "keyword": "nfl"}
{"mode": "get_growth", "source": "tiktok, google search", "keyword": "nfl", "percent_growth": ["30D"]}
{"mode": "get_top_trends", "type": "TikTok Trending Hashtags", "country": "GB", "limit": 5}
The MCP equivalents are get_top_trends(type="TikTok Trending Hashtags"), get_growth(keyword="nfl", source="tiktok", percent_growth=["7D", "1W", "14D", "30D", "1M"]), and get_time_series(keyword="nfl", source="tiktok"). REST top_trends with type TikTok Trending Hashtags and limit 3 on September 8, 2026 returned the same snapshot timestamp as the MCP omit-limit pull, with count 3: tiktokshop, movie, dance.
A hash prefix is stripped. Keyword #nfl still returned search_term nfl and the same 7D row as the bare token: 57.8 versus 85.3, growth -32.24%, volume 154,413,688. Pass the board label without #.
1W matched 7D exactly on nfl. 1M matched 30D exactly: 57.8 on 2026-09-05 versus 69.8 on 2026-08-07, growth -17.19%, volume 154,413,688 from 162,341,733. 14D is the window that flipped the sign, +35.36% from 42.7 on 2026-08-22 (volume 144,478,495). Extra windows still sit on one source-plus-keyword request.
The comma-separated nfl call returned sources_requested ["tiktok", "google search"], sources_successful 2, and trend_analysis.uptrend_ratio 1/2. TikTok 30D was -17.19% with 30 daily points. Google Search 30D was +46.15% (13.0 to 19.0) on recent_date 2026-09-05 with 261 weekly points, volume_estimated true, and estimated volume 349,000,000. Same preset, opposite signs, different grain. Read source_results before joining those percentages.
get_time_series for nfl on tiktok returned 30 daily rows from 2026-08-07 (index 69.8, volume 162,341,733) to 2026-09-05 (index 57.8, volume 154,413,688). Each row carried date, value, volume, keyword, and datatype tiktok. Index 100.0 landed on 2026-08-31 with volume 182,229,368. On 2026-08-12 the index printed 0.0 while volume stayed 116,298,730. The 0-100 column is not a rescaled volume.
paidpartnership shows why a 7D percent can look absurd. The series sat at 4.1 / 76,906,592 on 2026-08-29, 0.0 / 70,430,160 on 2026-08-28, then 100.0 / 229,422,842 on 2026-09-05. That is a one-day spike into the index peak, not a month of steady demand.
get_top_trends with type tiktok returned HTTP 400, error invalid_request, message Unknown type 'tiktok', plus 24 valid feed labels that include TikTok Trending Hashtags, TikTok Trending Hashtags by Category, and TikTok Trending Searches. get_growth with source TikTok Trending Hashtags and keyword mrbeast returned HTTP 400, error invalid_source, message 'tiktok trending hashtags' 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 tiktok.
The inverse fold exists on the feed name. type tiktok trending hashtags on get_top_trends returned HTTP 200 and echoed type TikTok Trending Hashtags, with the same omitted-country snapshot as the canonical label. llms.txt still describes feed labels as case-sensitive. The live handler accepted this lowercase TikTok string on September 8, 2026. Prefer the canonical TikTok Trending Hashtags spelling in stored configs.
get_time_series on tiktok, 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.
Omit percent_growth on nfl and both MCP and REST returned error growth_calculation_failed with message No growth window could be calculated. REST still used HTTP 200 with envelope statusCode 400. The inner row was period 12M, error date_out_of_range, data_start 2026-08-07, data_end 2026-09-05. Docs still present 12M as the default. That default needs a year of points. This source did not have them on this pull.
3M and YTD failed with the same date_out_of_range message. 12M failed the same way when it was requested explicitly on nfl, tiktokshopbacktoschool, lindsayclancy, and dollyparton. The working set on this grain is the short daily windows: 7D, 1W, 14D, 30D, 1M.
A 30D row that starts at index 0.0 still prints. tiktokshopbacktoschool, sohochic, fall, and dollyparton all returned 30D growth 999999.0 because baseline_value was 0.0 on 2026-08-07. dollyparton's matching volume row was 55,191,695 from 1,039,863, volume_growth +5207.59%. Quote volume_growth when the index baseline is zero. The 999999.0 print is the pipeline's answer to a zero index, not a demand figure to put in a forecast.
friends is the short-history percent that still has a non-zero baseline: 30D +2565.38% from 2.6 on 2026-08-07 to 69.3 on 2026-09-05, volume 76,278,768 from 65,926,530. Volume grew 15.7%. The index grew because 2.6 is a tiny baseline, not because volume grew 25 times.
country GB with limit 5 did not reuse the omitted-country snapshot. as_of_ts moved to 2026-09-08T06:04:49.819438+00:00, the payload added "country": "GB", and rank 1 became xyzabc. Rank 2 was backtoit. Rank 3 was school. Rank 4 was relateable. Rank 5 was onthisday. onthisday is the only overlap with the default top 5. TikTok Trending Hashtags is one of the feeds that accepts an ISO country code. Omit country for the default stored board. Pass GB when the question is Britain. The two lists are not a sort of the same 25 labels.
Passing type TikTok Trending Hashtags by Category with category Beauty and limit 5 returned a 5-deep Beauty chart, same omitted-country as_of_ts, rank 1 makeup, then lashes, skincareroutine, acne, lorealparispartner. That aisle list is a depth chart inside one official category string.
Omitting category still returned HTTP 200, count 25, same stamp. Every row was rank 1, each tagged with a different aisle: Beauty and Personal Care (makeup), Household Products (squishy), Food and Beverage (foodie), Education (senioryear), Science (physics). The mixed call is one winner per aisle, not a 25-deep Beauty page. Official names include Beauty, Dance, and Science. Skip category only when the job is "who leads each aisle," then filter in code.
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 TikTok hashtags vs searches page is the two-board follow-up. This split stays inside one feed: keyword volume versus live hashtag rank.
FAQ