Claude can pull live Polymarket trend data in one conversation through TrendsMCP: the Polymarket Top Markets board, the odds and volume on any open market, and five years of Google Search interest in Polymarket itself. Four tool calls do the job: get_top_trends for the board, get_context with the Polymarket source for odds and volume, get_time_series for history, and get_growth for the change over any window. The November 18 run found the 2028 Democratic presidential nominee market at the top of the board, with $1.33 billion traded in total and $3.37 million in the last 24 hours. Search interest in Polymarket rose 20 percent over 12 months, then fell from 58 to 24 in two weeks. The Free plan covers the whole pull with 100 requests a month.
The November 18 board was led by the 2028 Democratic nominee market
The Polymarket Top Markets board ranks prediction markets by the most money traded in the last 24 hours, and it updates live. The snapshot pulled on November 18 at 09:01 UTC looked like this:
| Rank | Market |
|---|---|
| 1 | Democratic Presidential Nominee 2028 |
| 2 | Jensen Huang rejoins NVIDIA by...? |
| 3 | Fed Decision in December? |
| 4 | Next CEO of NVIDIA? |
| 5 | Anthropic IPO Closing Market Cap |
| 6 | Republican Presidential Nominee 2028 |
| 7 | Israel x Iran ceasefire continues through...? |
| 8 | Next French Presidential Election |
| 9 | US x Iran Effective Ceasefire begins by...? (2 week pause) |
| 10 | What price will Bitcoin hit in November? |
Politics led the board. Two of the top ten markets were about the 2028 US presidential nominees, and a third was the December Fed decision. The rest split between tech leadership, the Anthropic IPO, geopolitics, and the November Bitcoin price. A board this top heavy in politics is a snapshot, not a permanent shape: the ranking follows 24 hour trading volume, so a single news event can reorder it by the next morning.
Four tool calls pull the whole picture in Claude
One prompt runs the whole pull once Claude is connected to TrendsMCP. The setup for Claude takes a custom connector URL and no API key. The prompt is:
Using Trends MCP, show me the Polymarket Top Markets board, the odds and volume on the top market, and the Google Search trend for "polymarket" over the past 5 years with 3 month and 12 month growth.
Claude turns that prompt into four tool calls. Each one is worth knowing by name, because the same four cover most trend questions on the sources in the catalog.
get_top_trends takes a board type and returns the ranked board. The call for this piece was get_top_trends with type Polymarket Top Markets and limit 10. It returned 10 rows and the snapshot timestamp, 2026-11-18T09:01:41Z. Cost: 1 request per 10 rows on MCP, so this call cost 1 request.
get_context takes a keyword and a source, and returns what people are posting, reading and betting about it right now. The call for this piece was get_context with keyword Democratic Presidential Nominee 2028 and source Polymarket. It returned 4 markets with odds, volume and 24 hour change. Cost: 1 request per source.
get_time_series takes a keyword and a source, and returns about five years of weekly history as a 0 to 100 index. The call for this piece was get_time_series with keyword polymarket and source Google Search. It returned 261 weekly points, from November 20, 2021 to November 14, 2026. Cost: 1 request per source.
get_growth takes a keyword, a source and one or more windows, and returns point to point growth for each window in a single call. The call for this piece was get_growth with keyword polymarket, source Google Search, and windows 3M and 12M. Cost: 1 request per source, every window included.
The same calls work over REST with a POST to https://api.trendsmcp.ai/api and a JSON body such as {"mode": "get_top_trends", "type": "Polymarket Top Markets", "limit": 10}. The data catalog lists all 54 data sources and the exact cost of each tool.
Search interest in Polymarket rose 20 percent in a year, then dropped by more than half in two weeks
Google Search interest in "polymarket" stood at 24 on the week of November 14, 2026, on a 0 to 100 index. That is up 20 percent from 20 a year earlier, on November 15, 2025, and up 14.29 percent from 21 three months earlier, on August 15, 2026. Both growth figures come from the get_growth call run for this piece.
The year in between was not a steady climb. Search interest peaked at 77 in the week of June 13, 2026, then slid through the summer to 18 in late August. It recovered through October, hit 58 in the week of October 31, then fell to 44 and 24 in the two weeks after. A 20 percent annual gain with a 59 percent drop in a fortnight is the shape of attention that spikes on events and drains fast. The latest weekly point is November 14, so the series lags the board by four days. Read the two together: the board says what traders are betting on today, the series says how much the wider public is searching for the platform at all.
Search demand for the exact phrase behind this piece is thin. Google Ads Keyword Planner returned too little volume to report for "polymarket trend data", 880 monthly searches for "polymarket data", and 18,100 for "polymarket api", both with low competition. Readers looking for this data search for the API or the data in general, not for a trend data phrase.
The top market carried $1.33 billion in volume
The get_context call on the top market returned four related markets. The main event, Democratic Presidential Nominee 2028, had $1,326,472,262 in total volume and $3,374,464 traded in the last 24 hours. One of the other markets covered the nominee's age bracket.
The call also returned the five sub-markets inside the main event, and they carry a warning about reading a slice as the whole field. All five were long shot names, and none had Yes odds above 0.005:
| Sub-market | Yes odds | Volume |
|---|---|---|
| Will Oprah Winfrey win the 2028 Democratic presidential nomination? | 0.001 | $54,672,982 |
| Will Bernie Sanders win the 2028 Democratic presidential nomination? | 0.003 | $52,995,342 |
| Will Chelsea Clinton win the 2028 Democratic presidential nomination? | 0.005 | $51,138,560 |
| Will Andrew Yang win the 2028 Democratic presidential nomination? | 0.001 | $49,284,155 |
| Will Hunter Biden win the 2028 Democratic presidential nomination? | 0.002 | $45,093,412 |
The five sub-markets sum to about $253 million, against $1.33 billion for the event as a whole, so most of the money sits in sub-markets this call did not return. Treat the odds above as the long tail, not the leaderboard. The volume figures are the solid part: they show where money actually moved, and the 24 hour figure of $3.37 million shows the market was still trading actively on November 18.
The age bracket sub-market added one more usable detail. Traders put the chance the nominee is under 45 at 0.375, 45 to 54 at 0.37, and 55 to 64 at 0.285, with the under 45 bracket down 0.015 in the last 24 hours. Odds here are prices, not polls: they say what traders will pay, not what voters will do.
The news around Polymarket this week was mostly promo codes
A get_context call with source News and keyword "polymarket" returned 10 stories from the past 7 days. Six of the ten were promo code and referral offers. The rest of the week had more substance: Politico reported on November 16 that prediction markets were feeling heat from the Trump administration, UFC.com covered Dana White on Polymarket UFC 334 on November 15, Axios reported on November 18 that Donald Trump Jr. explained the Kalshi split, and The Defiant reported the same day that Polymarket moved hourly and daily crypto bets to Chainlink.
The mix matters for anyone reading Polymarket trend data as a signal. Most of the news volume around the brand right now is affiliate marketing, not market news. The board and the odds carry the signal; the news count mostly carries the noise.
The full pull costs 4 requests on the Free plan
The four calls in this piece cost 4 requests in total: 1 for the 10 row board, 1 for the Polymarket context, 1 for the time series, and 1 for growth across both windows. The Free plan gives 100 requests a month with no credit card, so this whole workflow runs 25 times a month before hitting the limit. Pricing lists the paid tiers: Starter at $16 a month billed annually with 1,000 requests, Pro at $39 with 5,000, and Business at $159 with 25,000.
One limit matters for Free users. Boards on the Free plan lag by 24 hours unless the account holds purchased credits, while Starter, Pro and Business see them in real time. The snapshot in this piece was taken at 09:01 UTC on November 18 with a key that sees boards in real time, so a Free reader repeating the same call would see yesterday's board instead. Failed calls are free on every plan, so a mistyped board name costs nothing.
The best MCP server for live trend data compares TrendsMCP with the alternatives when the choice is still open. Once connected, the workflow above is the same every morning: board first, context on the leader, series and growth for the direction of travel.
FAQ
Does the Free plan show live Polymarket boards?
No. Free plan boards lag by 24 hours unless the account holds purchased credits. The history, odds and volume calls are the same on every plan.
Can Claude place bets on Polymarket?
No. Claude can read markets, odds, volume and search interest, and it cannot trade.
Why did the context call return long shot names instead of the leading candidates?
The call returned five sub-markets out of a much larger field. They summed to about $253 million of the event's $1.33 billion in volume. Ask for a specific candidate by name to get that candidate's market and odds.
