The Google Trends website has no public API and returns relative interest only. Trends MCP is the structured alternative: 0-100 index, absolute volume estimates, 5-year history, and live feeds via MCP.
Live data as of 2026-08-19 · Updated 2026-08-19
A Google Trends alternative for AI agents is a structured API, not a website wrapper. Trends MCP returns Google Search interest as a 0-100 index with about 5 years of weekly history, an absolute volume estimate where available, and live trending-search feeds. As of August 15, 2026, artificial intelligence had an index value of 46 and an estimated weekly volume of 717,000. The free plan includes 100 requests per month. Starter is $19 per month.
Artificial intelligence had an index value of 46 on August 15, 2026, down 49.45% from 91 on May 16, 2026 and down 17.86% from 56 on August 16, 2025. Estimated weekly volume moved from 1,420,000 on May 16, 2026 to 717,000 on August 15, 2026, against 873,000 a year earlier.
Electric vehicles had an index value of 39, down 60.61% over 3M and down 41.79% over 1Y, with estimated volume at 555,000 against 1,410,000 on May 16, 2026. Model context protocol had an index value of 7, down 88.89% over 3M and down 93.0% over 1Y from a peak of 100. OpenAI had an index value of 16, down 79.75% over 3M and down 78.67% over 1Y. AI agents had an index value of 31, down 68.04% over 3M and down 31.11% over 1Y.
The Google Trends website shows relative interest, related queries, and regional heat for a human clicking around. There is no official public API. Agents need a stable JSON object they can difference, store, and compare. Trends MCP maps that job onto 3 tools: get_trends for history, get_growth for period percentages plus volume, and get_top_trends for live breakout lists.
The source key is google search. That series is search-volume interest, not Search Console clicks, not Ads keyword planner bids, and not Discover feed impressions. A phrase can trend on Google while the team's site ranks nowhere.
get_top_trends with type Google Trends returns the current breakout list with no keyword. On August 19, 2026, the top 10 were hayden panettiere, inde navarrette, powerball winner, natalie harp, egg recall, hurricane lala, rebel creamery ice cream lawsuit, karoline leavitt, what time is the solar eclipse on august 12, and frank beard.
Discovery workflows start from that feed, then pass surviving phrases into get_growth for scored history. That 2-step pattern replaces the related-queries rabbit hole on the website, which is useful for humans and expensive for agents if it turns into hundreds of speculative phrases.
pytrends is a Python library that scrapes the Google Trends website. It returns relative 0-100 scores, breaks when the frontend changes, and inherits IP rate limits. It is fine for a laptop experiment and a poor dependency inside a production agent. The pytrends alternative page covers that path in more depth.
SerpApi, Bright Data, and similar SERP vendors can return Trends widgets as one of many scrape targets. Teams should check current vendor pricing for those platforms. They typically sell Google results broadly. They do not, by themselves, give one MCP session covering 25+ consumer-demand sources.
Content teams batch a keyword list every Monday and keep phrases with positive 3M and 1Y growth. Equity researchers watch issuer names into events, then confirm on news sentiment. Ecommerce teams compare Google Search with Amazon on the same phrase to separate research from shopping.
A practical batch looks like this. Store 30 phrases. Once a week, call get_growth with 3M and 1Y on google search. Keep phrases where both windows are positive. For the keepers, call get_trends weekly and archive the JSON. That pattern stays inside 100 requests per month for a small desk and scales on Starter at $19 per month when the list grows.
Low-volume phrases return weak quality scores or misses. The 0-100 index is relative to that keyword's own peak in the window, which is why two keywords at 80 are not equal in absolute searches. Volume estimates should not be pasted into a media plan as if they were Keyword Planner.
Quota is 100 requests per month on the free plan, Starter $19 per month. A crawl of 10,000 keywords overnight is a different product. Deduplicating the list and caching weekly snapshots is how research teams stay inside the month.
Sign up on the free tier, copy the API key, and make the first call with any HTTP client. The setup guide covers Claude, Cursor, and plain REST in 4 copy-paste blocks. The Google Search data page covers the source in more depth, and the Google Trends API page covers the API contract.
Tools for this workflow
get_trendsRetrieve about 5 years of weekly Google Search volume for any keyword as structured JSON.
get_trends(keyword='artificial intelligence', source='google search', data_mode='weekly')get_growthCompute period-over-period Google Search growth with absolute volume estimates.
get_growth(keyword='artificial intelligence', source='google search', percent_growth=['3M', '1Y'])get_top_trendsPull live Google Trends breakout topics in real time via API.
get_top_trends(type='Google Trends', limit=25)FAQ