A GnuCOBOL job talks to Trends API with one POST to https://api.trendsapi.ai/api, a Bearer key, and a JSON body that names a mode. There is no official GnuCOBOL library. CALL "SYSTEM" wraps curl. jq parses JSON. The first decode reads the envelope, then a second decode reads the body string. Pulled on 2026-11-14, Google Search for cobol programming scored 19.0 for the week of 2026-11-07, versus 24.0, down 20.83%. gnu cobol scored 10.0 versus 0.0. One-word gnucobol scored 0.0. cobol compiler scored 15.0 versus 2.0, up 650.0%. YouTube for cobol programming scored 32.0 versus 6.0, the opposite 7D sign from Search. Do not copy cobol compiler 15.0 onto a COBOL adoption alert that only watches the phrase token at 19.0. A rank-3 phrase whose weekly Search printed 2.0 while daily Search printed 100 is documented in Michigan Senate Race live rank versus weekly Search.

How does a GnuCOBOL client call Trends API?

POST https://api.trendsapi.ai/api
Authorization: Bearer <api_key>
Content-Type: application/json

{"mode":"get_growth","source":"google search","keyword":"cobol programming","percent_growth":["7D","30D","3M","12M"]}
>>SOURCE FORMAT FREE
identification division.
program-id. trends-growth.

environment division.

data division.
working-storage section.
01  cmd pic x(1024).
01  rc  pic s9(9) binary.

procedure division.
    string
        "curl -sS -X POST https://api.trendsapi.ai/api "
        "-H 'Authorization: Bearer "
        function getenv("TRENDSAPI_API_KEY")
        "' -H 'Content-Type: application/json' "
        "-d '{""mode"":""get_growth"",""source"":""google search"","
        """keyword"":""cobol programming"","
        """percent_growth"":[""7D"",""30D"",""3M"",""12M""]}' "
        "| jq -r .body | jq ."
        into cmd
    end-string
    call "SYSTEM" using cmd returning rc
    goback.

CALL "SYSTEM" owns the HTTP client by wrapping curl. The first jq reads statusCode and body. jq -r .body prints the inner string. A second jq on that string is required before reading recent_value. FUNCTION GETENV reads the Bearer key. The units ship outside this API:

https://gnucobol.sourceforge.io
https://curl.se/docs/manpage.html
https://jqlang.github.io/jq
https://www.ibm.com/docs/en/cobol-zos/6.4?topic=statements-json-parse-statement

curl and jq are client tools. They are not Trends API sources. There is no COBOL package-registry download-count source on this API. An IBM Enterprise COBOL job can send the same POST and parse with JSON PARSE. That runtime is not a Trends source either.

The response envelope is {"statusCode": int, "body": string}. body is a JSON string and must be parsed a second time before reading recent_value. A Fortran client that does the same double parse with curl and json-fortran is on Fortran Trends API. An Ada client that does the same double parse with AWS.Client is on Ada Trends API.

What did get_growth return for COBOL language tokens?

get_growth for cobol programming on google search tips on 2026-11-07. Store the exact strings. cobol language is a different series. Two-word gnu cobol closed 9 points below programming. One-word gnucobol closed at 0.0.

Keyword Period Recent date Recent Baseline date Baseline Growth
cobol programming 7D 2026-11-07 19.0 2026-10-31 24.0 -20.83%
cobol programming 14D 2026-11-07 19.0 2026-10-24 34.0 -44.12%
cobol programming 30D 2026-11-07 19.0 2026-10-10 21.0 -9.52%
cobol programming 3M 2026-11-07 19.0 2026-08-08 25.0 -24.0%
cobol programming 12M 2026-11-07 19.0 2025-11-08 27.0 -29.63%
cobol programming YTD 2026-11-07 19.0 2026-01-03 21.0 -9.52%
cobol language 7D 2026-11-07 17.0 2026-10-31 21.0 -19.05%
cobol language 14D 2026-11-07 17.0 2026-10-24 24.0 -29.17%
cobol language 12M 2026-11-07 17.0 2025-11-08 23.0 -26.09%
cobol 7D 2026-11-07 15.0 2026-10-31 17.0 -11.76%
cobol YTD 2026-11-07 15.0 2026-01-03 12.0 +25.0%
gnu cobol 7D 2026-11-07 10.0 2026-10-31 0.0 vs 0.0
gnucobol 7D 2026-11-07 0.0 2026-10-31 0.0 0.0%
gnucobol 3M 2026-11-07 0.0 2026-08-08 27.0 -100.0%
ibm cobol 7D 2026-11-07 2.0 2026-10-31 2.0 0.0%
ibm cobol 3M 2026-11-07 2.0 2026-08-08 14.0 -85.71%
micro focus cobol 7D 2026-11-07 5.0 2026-10-31 0.0 vs 0.0
micro focus cobol 3M 2026-11-07 5.0 2026-08-08 32.0 -84.38%
cobol compiler 7D 2026-11-07 15.0 2026-10-31 2.0 +650.0%
cobol compiler 14D 2026-11-07 15.0 2026-10-24 29.0 -48.28%
opencobol 7D 2026-11-07 0.0 2026-10-31 0.0 0.0%

get_time_series for cobol programming on Google Search closed at 19.0 on 2026-11-07, matching get_growth 19.0. Adjacent weeks: 24.0 on 2026-10-31, 34.0 on 2026-10-24, 20.0 on 2026-10-17, 21.0 on 2026-10-10. Peak in this pull is 100.0 on 2026-02-28. Next highest points are 92.0 on 2025-02-22, then 66.0 on 2025-02-15, then 64.0 on 2023-10-21. Tip 19.0 is not that February 2026 high. 12M printed 19.0 versus 27.0 (-29.63%). Same Saturday tip, opposite story depending on the window: 7D is -20.83% while 14D is -44.12%.

cobol language closed at 17.0 versus 21.0, down 19.05%. Same 7D sign as programming. Adjacent weeks: 21.0 on 2026-10-31, 24.0 on 2026-10-24, 23.0 on 2026-10-17, 18.0 on 2026-10-10. Peak in the language series is 100.0 on 2025-02-22, then 99.0 on 2026-02-28. Programming peaked in 2026. Language peaked in 2025. On 2026-02-28 programming printed 100.0 while language printed 99.0. Matching Saturday tips are not matching series. Do not copy programming 19.0 onto language 17.0.

Bare cobol closed at 15.0, down 11.76% from 17.0. YTD printed 15.0 versus 12.0 (+25.0%). Programming YTD printed 19.0 versus 21.0 (-9.52%). Same Saturday tip, opposite YTD signs. Peak in the bare series is 100.0 on 2026-02-28, matching the programming peak week. Do not treat Search 15.0 as language adoption.

gnu cobol closed at 10.0 versus 0.0 on 7D, 14D, 30D, 3M, 12M, and YTD. Report 10.0 versus 0.0. The payload marked data_quality low. Adjacent weeks: 0.0 on 2026-10-31, 0.0 on 2026-10-24, 0.0 on 2026-10-17, 0.0 on 2026-10-10. Peak in this pull is 100.0 on 2026-09-26, then 95.0 on 2026-07-18, then 77.0 on 2026-09-05. The two-word compiler phrase sat 9 points below cobol programming 19.0. One-word gnucobol closed at 0.0 versus 0.0 on 7D. 3M printed 0.0 versus 27.0 (-100.0%). Adjacent weeks: 0.0 on 2026-10-31, 0.0 on 2026-10-24, 0.0 on 2026-10-17, 1.0 on 2026-10-10. Peak in the one-word series is 100.0 on 2026-09-05 and 2026-09-26, then 78.0 on 2026-09-12. Two-word close 10.0 is not one-word close 0.0. Do not copy gnu cobol 10.0 onto gnucobol 0.0.

ibm cobol closed at 2.0 versus 2.0 on 7D. 3M printed 2.0 versus 14.0 (-85.71%). 12M printed 2.0 versus 4.0 (-50.0%). micro focus cobol closed at 5.0 versus 0.0 on 7D. Report 5.0 versus 0.0. 3M printed 5.0 versus 32.0 (-84.38%). Do not copy micro focus cobol 5.0 onto gnu cobol 10.0.

cobol compiler closed at 15.0 versus 2.0, up 650.0%. Opposite 7D sign from programming. 14D printed 15.0 versus 29.0 (-48.28%). A helper that only stores 7D then reports a 650.0% climb off a 2.0 floor while 14D is already -48.28%. 3M printed 15.0 versus 51.0 (-70.59%). Do not copy cobol compiler 15.0 onto cobol programming 19.0.

opencobol printed 0.0 versus 0.0 on every window. The payload marked data_quality low. Store the zeros. OpenCOBOL was the prior name of GnuCOBOL. The retired string is not the compiler series.

Neighbor tools on the same pull, same source, same 7D window, same 2026-11-07 tip:

Keyword 7D recent 7D baseline 7D growth 12M recent 12M baseline 12M growth
fortran programming 28.0 31.0 -9.68% 28.0 54.0 -48.15%
pascal programming 26.0 31.0 -16.13% 26.0 66.0 -60.61%
cobol programming 19.0 24.0 -20.83% 19.0 27.0 -29.63%
gnu cobol 10.0 0.0 vs 0.0 10.0 0.0 vs 0.0
ada programming 12.0 12.0 0.0% 12.0 32.0 -62.5%
cobol compiler 15.0 2.0 +650.0% 15.0 31.0 -51.61%
gnucobol 0.0 0.0 0.0% 0.0 0.0 0.0%

fortran programming 28.0 sat 9 points above cobol programming 19.0. pascal programming 26.0 sat 7 points above. ada programming 12.0 sat 7 points below. cobol compiler printed +650.0% on 7D while cobol programming printed -20.83%. Do not copy fortran programming 28.0 onto cobol programming 19.0. High-volume php in this session tipped on 2026-11-07 at 25.0 versus 38.0 with estimated volume 556,000. That is the same Saturday tip as the COBOL keywords, not the older 2026-10-31 tip that still held people-keywords. Store the session pull and the exact recent_date per keyword. A Pascal Trends API helper stores a language neighbor on the same 7D window. A Fortran Trends API helper stores another.

What did YouTube, News, and Wikipedia return?

YouTube for cobol programming printed the opposite 7D sign from Search. Keyword Google News for the programming phrase printed zeros on 7D. Wikipedia titles resolve on their own. Compare Wikipedia on volume.

YouTube for cobol programming tipped 32.0 on 2026-11-07 versus 6.0, up 433.33%, matching get_time_series 32.0, while Search sat at 19.0 down 20.83%. Opposite 7D signs. Store both. Adjacent weeks: 6.0 on 2026-10-31, 39.0 on 2026-10-24, 42.0 on 2026-10-17, 0.0 on 2026-10-10. 14D printed 32.0 versus 39.0 (-17.95%), the same 14D sign as Search. 30D, 3M, 12M, and YTD printed versus 0.0 (growth 999999). Report 32.0 versus 0.0. Peak in the YouTube series is 100.0 on 2026-02-28. Next highest points are 76.0 on 2022-06-18, then 71.0 on 2024-07-06, then 70.0 on 2024-05-04 and 2025-02-22. On 2026-02-28 Search printed 100.0 while YouTube printed 100.0. Those are matching weeks. Do not treat YouTube 32.0 as language adoption while Search sits at 19.0 after a 100.0 close in February 2026.

YouTube for cobol language closed at 19.0 versus 0.0 on 7D, 14D, 30D, 3M, 12M, and YTD. Report 19.0 versus 0.0. The payload marked data_quality low. Language YouTube 19.0 is not programming YouTube 32.0.

YouTube for gnu cobol printed 0.0 versus 0.0 on every window, with data_quality low. Store the zeros. gnu cobol YouTube 0.0 is not gnu cobol Search 10.0.

Keyword Google News for cobol programming printed 0.0 versus 0.0 on every window, with data_quality low. The first News call in this session returned 502 source_unavailable. The retry printed zeros. Store the miss and the zeros. Do not fill News from Search 19.0. Keyword News and the Google News Top News board are different feeds. Mode docs sit at https://trendsapi.ai/docs.

Wikipedia for title COBOL resolved. October 2026 scored 11.8 with 22,741 pageviews versus September 7.0 with 19,649 (volume +15.74%). 7D and 14D collapsed (collapsed_range: recent and baseline resolved to the same monthly point). 3M was 11.8 / 22,741 versus July 6.0 / 18,945 (volume +20.04%). 12M was 11.8 / 22,741 versus October 2025 13.9 / 24,164 (volume -5.89%). YTD was 11.8 / 22,741 versus December 2025 6.6 / 19,356 (volume +17.49%). Peak in this pull is 100.0 / 80,422 in February 2025, then 56.8 / 52,169 in February 2026, then 53.7 / 50,156 in March 2025. Title GnuCOBOL resolved. October scored 12.9 with 476 pageviews versus September 5.2 with 396 (volume +20.2%). 12M was 12.9 / 476 versus October 2025 31.3 / 667 (volume -28.64%). Peak is 100.0 / 1,381 in March 2025. The language page is about 47.8 times the GnuCOBOL volume. Title IBM COBOL resolved. October scored 2.2 with 300 pageviews versus September 2.4 with 303 (volume -0.99%). 12M was 2.2 / 300 versus October 2025 7.6 / 401 (volume -25.19%). IBM COBOL volume is about 0.63 times GnuCOBOL and about 0.013 times the language title. Title Micro Focus resolved. October scored 0.0 with 1,956 pageviews versus September 2.3 with 2,258 (volume -13.37%). Score 0.0 with 1,956 views is the finding. Compare on volume. Do not join 22,741 language views onto Search 19.0. Do not treat GnuCOBOL at 476 views as the language page. The language title is the page to store. Wikipedia source docs sit at https://trendsapi.ai/trends/wikipedia-trends.

There is no COBOL package-registry source on this API. npm covers npmjs. python covers PyPI. A COBOL watchlist that wants compiler downloads has no third source to join onto Search 19.0.

How does a watchlist store live boards next to weekly scores?

get_top_trends has no keyword. Pair the live Google Trends board with get_growth on the exact board string. Live rank, 1-day rank_change, and 7-day rank_change are three feeds. The rank snapshot can stamp at the same second as the mover snapshot, or days earlier than Wikipedia Trending.

POST https://api.trendsapi.ai/api
Authorization: Bearer <api_key>
Content-Type: application/json

{"mode":"get_top_trends","type":"Google Trends","limit":25}

On 2026-11-04T07:01:36Z the live Google Trends board ranked michigan senate race at 3, with no row on the 1-day mover list. Wikipedia Trending, YouTube Trending, and Google News Top News stamped 2026-11-14T09:01:40Z, ten days later. Weekly Search for michigan senate race scored 2.0 versus 2.0 on the 2026-10-31 tip, flat. Daily Search printed 100 on 2026-11-04, then 0 on 2026-11-13. YouTube scored 100.0 versus 7.0 on the 2026-11-07 tip. Wikipedia Trending omitted the title on the current stamp. The helper that stores those rows is the blog in this batch: Michigan Senate Race live rank versus weekly Search. Do not copy that daily 100 onto cobol programming 19.0. Board-shape notes for Google Trends live feeds sit at https://trendsapi.ai/trends/google-trends.

Free tier is 100 successful 200s per month. Only 200 responses count against quota. Five language tokens plus one live board is 6 calls. A Monday run that also pulls YouTube and Wikipedia for the programming phrase is 8 calls. Four weeks is 32 calls and still fits the free cap. Starter is 5,000, Pro 25,000, Business 100,000.

Envelope checklist

  1. POST https://api.trendsapi.ai/api with Authorization: Bearer <api_key>.
  2. Parse the transport JSON, then parse the body string a second time.
  3. Store cobol, cobol programming, cobol language, gnu cobol, gnucobol, ibm cobol, micro focus cobol, cobol compiler, and opencobol as separate series.
  4. Compare Wikipedia COBOL on volume at 22,741 October views, not GnuCOBOL at 476, not IBM COBOL at 300, and not Micro Focus at 1,956.
  5. Pair get_top_trends rank with both rank_change windows and get_growth on the exact board string. Store each keyword's recent_date.