Prediction Markets Scraper - Kalshi & Polymarket Odds

Prediction Markets Scraper collects live odds from Kalshi and Polymarket in one run. Each row has the yes/no price, best bid/ask, volume, liquidity, category and close time, with optional order book depth, price history, recent trades and top holders. Filter by category, minimum volume or liquidity and sort order, or switch on monitor mode to receive only new markets and price moves. It also matches the same question across both exchanges and reports the price gap. Export CSV, Excel, JSON or API. No login, no API key. Verified 2026-10-10: $0.02 per market row.

Prediction Markets Scraper reads the public APIs of Kalshi and Polymarket and returns one normalized schema for both. A real test run on 2026-10-07 returned 10 market rows plus 10 cross-platform matches in 12 seconds.

🎯 What does the Prediction Markets Scraper do?

• Scrapes Kalshi prediction market odds and Polymarket odds in the same run, or either one alone, normalized to one yesPrice and noPrice between 0 and 1.
• Filters by search terms, category, status, minimum volume and minimum liquidity (Polymarket), with six sort orders.
• Monitor mode (onlyNew) returns only markets earlier runs did not deliver, and can re-deliver a market when its Yes price moves by an amount you set. Skipped rows cost nothing.
• Adds order book depth, price history, recent trades, top holders (Polymarket) and wallet mode for up to 100 wallets when you turn them on.
• Finds the same question on Kalshi and Polymarket and outputs the price gap, the spread in percent and which exchange is cheaper.

📊 What data can you extract from Kalshi and Polymarket?

Field	Type	Platforms	Example
recordType, platform	string	both	market-summary, kalshi
marketId, eventId, eventTitle, outcomeLabel	string	both	KXOAIANTH-40-ANTH, KXOAIANTH-40
question, category, tags	string, array	both (tags Polymarket)	Will Open AI or Anthropic IPO first?
status, yesPrice, noPrice	string, number (0 to 1)	both	open, 0.94, 0.06
yesBid, yesAsk, noBid, noAsk, lastPrice	number	both (no* Kalshi only)	0.94, 0.95
volume, volume24h	number	both, when above zero	207250, 54
openInterest, liquidity, spread	number	Kalshi, Polymarket, both	54186.64, 63561
priceChange1h, 1d, 1w, 1m, 1y	number	Polymarket only	price change per window
previousClose, volume1w, volume1m, volume1y	number	previousClose Kalshi, volumes Polymarket	previous UTC day close, volume per window
resolved, result, rules, resolutionSource	mixed	both	settlement and rules text
changeType, previousYesPrice	string, number	monitor mode	new or price-move, and the earlier price
openTime, createdTime, closeTime	string	both	2040-01-01T04:59:00Z
orderBook, priceHistory, recentTrades, topHolders	object, array	when enabled	{"bids":[...],"asks":[...]}, [{"time":"...","price":0.026}]
wallet, size, avgPrice, currentValue, cashPnl, realizedPnl	mixed	wallet mode	one row per position, recordType wallet-position
url, scrapedAt	string	both	https://kalshi.com/markets/..., 2026-10-07T19:31:59.699Z

The dataset schema lists 77 typed fields; Polymarket rows also carry outcomes, outcomePrices, negRisk and conditionId. Columns the exchange does not provide are left out: no N/A, null or placeholder values.

🧾 Sample output

Real Polymarket row from a run on 2026-10-07 (default settings, no extras):

{
  "recordType": "market-summary",
  "platform": "polymarket",
  "marketId": "559651",
  "question": "Xi Jinping out before 2027?",
  "status": "active",
  "yesPrice": 0.0265,
  "noPrice": 0.9735,
  "lastPrice": 0.027,
  "volume": 14425353,
  "volume24h": 7491,
  "closeTime": "2027-01-01",
  "url": "https://polymarket.com/event/xi-jinping-out-before-2027",
  "scrapedAt": "2026-10-07T19:32:00.617Z"
}

Real cross-platform match from a run on 2026-10-07 (same question on both exchanges):

{
  "recordType": "cross-match-record",
  "question": "Will Gavin Newsom be the Democratic Presidential nominee in 2028?",
  "matchScore": 79,
  "kalshi": { "marketId": "KXPRESNOMD-28-GN", "yesPrice": 0.14 },
  "polymarket": { "marketId": "559652", "yesPrice": 0.1705 },
  "priceGap": -0.0305,
  "absPriceGap": 0.0305,
  "cheaperPlatform": "kalshi",
  "spreadPercent": 21.79,
  "scrapedAt": "2026-10-07T19:33:07.290Z"
}

The kalshi and polymarket objects also carry eventId, question and url (shortened here). priceGap is the Kalshi yes price minus the Polymarket yes price.

💰 How much does it cost to scrape prediction markets?

Pay per event. You are charged only for rows saved to your dataset. Rows with missing data, failed requests and rows skipped by monitor mode are never charged.

Event	What it is	Price (FREE plan)
market-summary	One market row	$0.02 each ($20.00 per 1,000)
cross-match-record	One matched pair with the price gap	$0.01 each
order-book-record, price-history-record, recent-trades-record, holders-record	Each extra, only when data came back	$0.01 each
wallet-position-record	One current Polymarket position of a wallet	$0.003 each ($3.00 per 1,000)

On higher Apify plans the market row price drops to $0.019, $0.018 or $0.017 depending on the plan; the extras keep the same price. Example: 10 markets with three extras plus 10 cross-matches = 10 x $0.02 + 40 x $0.01 = $0.60. A wallet run returning 40 positions costs 40 x $0.003 = $0.12. The default run (10 markets, no extras, cross-match off) costs $0.20; each cross-match found adds $0.01. A monitor run that finds nothing new returns 0 rows and charges $0. Free Apify accounts are limited to 10 market rows per run.

🚀 How to scrape Kalshi and Polymarket (3 steps)

1. Open the Actor in Apify Console and keep both platforms selected.
2. Optionally add search terms such as bitcoin, set Max items and switch on extras.
3. Click Start, then download CSV, Excel or JSON from the Output tab.

⚙️ Input

Field	Default	What it does
platforms	["kalshi","polymarket"]	Exchanges to read.
status	open	open, closed, settled, all.
searchTerms	[]	Keep only markets that match any phrase. Polymarket searches title, description and tags; Kalshi matches literally.
categories	[]	Keep only markets in these categories, for example crypto.
minVolume	0	Minimum volume. USD on Polymarket, contracts on Kalshi.
minLiquidity, liquidOnly	0, false	Minimum liquidity in USD (Polymarket only); drop zero-volume markets.
sortBy	default	default (most traded first), volume, volume24h, liquidity, newest or endingSoon.
maxPerPlatform	300	Markets fetched per exchange (up to 5,000).
maxItems	100	Maximum market-summary rows; cross-match rows are capped separately.
onlyNew	false	Monitor mode: skip markets an earlier run of the same search already delivered.
priceChangeThreshold	0	With onlyNew, re-deliver a seen market when its Yes price moved this much (0.05 is 5 cents).
stateKey, resetSeen	empty, false	Separate monitor memories for the same filters; clear memory before the run.
includeOrderBook, includePriceHistory, includeTrades	false	Add top 10 bids and asks, the hourly price series, the last 20 trades.
includeHolders	false	Add the top holders per outcome (Polymarket only), $0.01 per market.
includeCrossMatch	true	Add one row per same-question pair (needs both platforms).
crossMatchDepth	3000	Markets scanned per exchange to find pairs (paid accounts).
walletAddresses	[]	Wallet mode: current Polymarket positions of up to 100 wallets, $0.003 per position.

{
  "platforms": ["kalshi", "polymarket"],
  "status": "open",
  "categories": ["crypto"],
  "minVolume": 1000,
  "sortBy": "volume24h",
  "maxItems": 20,
  "onlyNew": true,
  "includeOrderBook": true,
  "includeCrossMatch": false
}

Rows alternate between exchanges, so a small maxItems gives you both. With includeCrossMatch on, a run can return up to maxItems market rows plus up to maxItems cross-match rows. Kalshi has no server side filter or sort, so sorting there works inside the first 1,500 markets read. Monitor mode keeps its memory in the key-value store prediction-markets-seen in your account, one record per search; rows are marked changeType: "new" or "price-move".

📦 Output

The dataset has three record types: market-summary, cross-match-record and wallet-position. In the Output tab you get a Markets view, a Cross-platform matches view and a Wallet positions view. Download as JSON, CSV, Excel, XML or HTML, or fetch it through the API. A search that matches nothing finishes as SUCCEEDED and charges nothing. If both exchanges cannot be read, the run fails with a clear message and charges nothing.

🆚 Prediction markets scraper vs alternatives

Store pages read on 2026-10-10. "Not stated" means the listing does not show the feature.

Feature	This Actor	scrapesage	chorelet
Price per 1,000 markets	$20 ($17 Gold)	from $1.66	from $1.40 ($2 on FREE)
Monitor, only new rows	Yes, plus price-move re-delivery	Yes, odds and volume moves	Not stated
Filters and sort	Category, volume, liquidity, 6 sort orders	Tags, volume, liquidity, sort	Tags, volume, 5 sort orders
Order book	Top 10 levels per market	Not stated	Top of book only
Holders and wallets	Holders and wallet positions, Polymarket	Holders, wallet P&L, trades	Not stated
Cross-platform price gap	Yes, per question	Yes, arbitrage mode	Not stated

Other Actors read the same day: scrapebench charges $0.0015 per market with a volume filter, and feedforge from $4.00 per 1,000 with search and minimum volume only. We are 6 to 13 times more expensive per plain market row, so if you only need a flat price dump, a cheaper Actor will do. scrapesage covers more modes, such as events, trade feeds and wallet P&L totals, and costs far less. The higher price here buys one schema for both exchanges, per-market order books, hourly history and same-question gaps in one run.

💡 Use cases

Arbitrage research (sort by absPriceGap), forecast dashboards for BI tools or LLM pipelines, backtesting with hourly priceHistory, and newsletters that cite odds with source URLs.

🔌 Run via API, schedule and integrations

curl -X POST "https://api.apify.com/v2/acts/SosMvBELkUIJNTIG7/runs?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"searchTerms":["federal reserve"],"sortBy":"volume24h","maxItems":50}'

Use the Apify API client for JavaScript or Python, schedules, and connect the dataset to Zapier, Make, n8n, Google Sheets or webhooks.

🤖 Prediction Markets Scraper MCP server for AI agents

AI agents can call the Prediction Markets Scraper through the Apify MCP server at https://mcp.apify.com?tools=RecordsData/prediction-markets-odds-scraper.

Add this to your MCP client configuration (Claude Desktop, Cursor, VS Code or any client that supports remote MCP servers):

{
  "mcpServers": {
    "apify": {
      "url": "https://mcp.apify.com?tools=RecordsData/prediction-markets-odds-scraper"
    }
  }
}

Example agent prompt:

Use the Prediction Markets Scraper to find open Kalshi and Polymarket markets about the federal reserve, keep cross-platform matches on, and tell me which question has the biggest price gap between the two exchanges. Run it again tomorrow with onlyNew on and list only the new markets.

Choose this tool when an agent needs live prediction market probabilities from Kalshi and Polymarket, or a same-question price comparison. It takes the same fields as the Input table above and returns market rows plus cross-match rows, with optional extras billed separately. Set onlyNew for new markets only, categories, minVolume and sortBy to narrow the list, and includeHolders or walletAddresses for Polymarket holders and wallet positions.

⚖️ Is it legal to scrape Kalshi and Polymarket?

The Actor reads public market data from the exchanges' public APIs. It does not log in or collect personal data. Review each exchange's terms for your own use case. This is an independent tool, not affiliated with Kalshi or Polymarket, and prices are not financial advice.

❓ Frequently asked questions

Why did I get 0 results from the Prediction Markets Scraper? Your search terms, categories, volume or liquidity filter, or market status matched no Kalshi or Polymarket markets. Remove the filters or set the status input to all and run the Prediction Markets Scraper again. A run that matches nothing, or a monitor run with nothing new, still finishes as SUCCEEDED and is not charged.

Do I need a Kalshi or Polymarket account or API key? No. Kalshi and Polymarket both publish market data openly, and the Prediction Markets Scraper reads those public APIs. You need no exchange account, login, cookies, API key or proxy, only an Apify account to run it, call the API or schedule snapshots.

Why are the prices different on Kalshi and Polymarket? They are separate order books with their own traders, so the same question can trade at different prices. Kalshi yesPrice is the last price, falling back to the bid/ask midpoint, while Polymarket yesPrice is the bid/ask midpoint, falling back to the outcome price. Both are normalized to a number between 0 and 1.

How many Kalshi and Polymarket markets can I get per run? The maxPerPlatform input accepts up to 5,000 markets per exchange per run, default 300, within a 40 second discovery budget. The maxItems input caps the market rows saved, default 100. Free Apify accounts get 10 market rows per run, and paid accounts can raise the limits.

How does cross-platform matching work in the Prediction Markets Scraper? It compares question keywords and close dates across Kalshi and Polymarket, and keeps the best candidate with a matchScore. A deeper crossMatchDepth finds more pairs. Matching is heuristic, so check the two question texts before you trade on a price gap.

Why are some columns missing on some Prediction Markets Scraper rows? Rows only contain what the exchange returned, with no N/A, null or placeholder values. For example openInterest and the noBid and noAsk fields exist for Kalshi but not for Polymarket, and volume is left out when it is zero. Liquidity and the price changes exist only on Polymarket rows.

Am I charged for failed requests or empty rows on the Prediction Markets Scraper? No. It charges only for saved market rows, cross-match rows, wallet positions and extras (order book, price history, recent trades, top holders) that returned real data. Failed requests, rows with missing data and rows that monitor mode skips are not billed, and the run stops cleanly at the maximum charge limit you set.

How fresh is the Kalshi and Polymarket odds data? The Prediction Markets Scraper reads live prices from the Kalshi and Polymarket public APIs at run time, and every row carries a scrapedAt timestamp. Schedule the actor in Apify to refresh as often as you need, for example every hour.

Does the Prediction Markets Scraper return closed or settled markets? Yes. The default status is open, but you can set the status input to closed, settled or all, and the Prediction Markets Scraper returns those Kalshi and Polymarket markets with the same normalized schema, so open and settled rows can be compared in one dataset.

🔗 Other RecordsData scrapers

SEC EDGAR Filings Scraper, DefiLlama TVL Scraper and Exchange Rates Scraper.

🆘 Support

Questions? Use the Issues tab or write to contact.punkrecordsdata@gmail.com.

Last updated: 2026-10-11