Polymarket & Kalshi Scraper – Odds, Volume & Price History

Polymarket and Kalshi scraper (plus Manifold) in one unified format: odds and probabilities, YES/NO prices, outcomes, volume, 24h volume, liquidity, open interest, close time, resolution and optional price history. Search by keyword or category, open or resolved markets. Official public APIs.

Use it instead of a single-venue scraper when you need Kalshi and Polymarket side by side, or, in weather mode, when you need weather markets joined to the station they settle on and the observed temperature or rain at that station.

Price: $1 per 1,000 markets. Weather station data is included at no extra cost. Price history is an optional add-on at $2 per 1,000 markets. Failed venues and empty searches are never charged.

What does Prediction Markets Scraper do?

It calls the official public APIs of the three biggest prediction market venues (no login, no API keys, no browser) and turns their very different data models into one flat schema:

Venue	Data source	What you get
Polymarket	Gamma API (events, markets, search) + CLOB prices-history	Every market inside an event, outcome prices, bid/ask, volume, 24h volume, liquidity, token IDs, condition IDs
Kalshi	Trade API v2 (events, markets, series, candlesticks)	Market tickers, YES/NO prices, bid/ask, volume, 24h volume, open interest, rules, settlement result
Manifold	API v0 (search-markets, market, bets)	Binary, multiple-choice and numeric markets, probabilities, volume, liquidity, bettor counts, resolution

For every market you get:

• 🎯 Odds: yesPrice, noPrice and outcomes[{name, price}] (0–1, the implied probability)
• 💵 Activity: volume, volume24h, liquidity, openInterest, bestBid, bestAsk, lastPrice
• 🗓 Lifecycle: status (open / upcoming / closed / resolved), openTime, closeTime, resolution
• 🏷 Context: question, eventTitle, subtitle (candidate or bucket), category, tags, resolution rules (description), market url
• 📈 Price history (optional): priceHistory: [{time, price}], hourly for ranges up to 1 month and daily for longer ranges
• 🧩 Raw venue fields under extra, e.g. Polymarket clobTokenIds / conditionId, Kalshi seriesTicker, Manifold outcomeType / uniqueBettorCount

Weather mode: Kalshi & Polymarket weather markets with station data

Set mode to weather to get Kalshi "Climate and Weather" markets and Polymarket weather markets (daily highest/lowest temperature in 40+ cities, rain, snow). Kalshi weather series are discovered automatically, so new cities appear without an update. Each market row also gets:

• 📍 Settlement station: stationId (ICAO, e.g. KNYC, KLGA, EGLC, RKSI), NWS climate site (cliStation, from CLINYC-style rules), name, coordinates, elevation and time zone. The station is read from the market's own rules or resolution source.
• 🌡 Observed values at that station: observedValue (the high, low or precipitation the market settles on), observedHigh, observedLow, observedPrecipIn, observedSnowIn.
• ✅ Bucket check: bucketLow / bucketHigh / unit parsed from the outcome label, and observedInBucket, which says whether the observed value lands in this bucket. While the day is still running, it is false once the bucket can no longer win and null while it can.
• 🧾 Kalshi settled value (settledValue) next to the observation, so you can audit settlements.

Where the observations come from:

Market	Observation source	observedStatus
Kalshi US daily high/low (rules name a CLIxxx site)	Official NWS Daily Climate Report (CLI), archived by the Iowa Environmental Mesonet	final the morning after the date
Polymarket temperature markets (NOAA time series / Wunderground station) and any day still in progress	METAR/ASOS reports of the airport station, local calendar day	complete or in-progress
Polymarket Hong Kong	Hong Kong Observatory, since-midnight max/min (today only)	in-progress
Future dates	none yet	pending

Live check (8 Oct 2026, 400 resolved weather markets: 200 Kalshi, 200 Polymarket): observedInBucket matched the venue's resolution in 400 of 400 markets, and the NWS value equalled Kalshi's settled value in all 200 Kalshi markets.

Why use it?

• One schema for three venues. Compare the same question across Polymarket, Kalshi and Manifold without writing three integrations.
• Reliable. It uses official JSON APIs with retries and backoff and does no HTML scraping. One venue failing never fails the run.
• Cheap and predictable. $0.001 per market, a fraction of single-venue actors. maxItems and Apify's Max cost per run cap your spend exactly.
• Agent-friendly. Calling it with {} returns the 30 most active markets in about 10 seconds. The input is small and typed, and the output is flat.

Use cases

• Trading and arbitrage research: spot price differences for the same event across venues.
• News, politics and macro dashboards: track election, Fed and CPI odds on a schedule.
• Quant and data science: build datasets of resolved markets with their price paths for calibration studies.
• AI agents and RAG: give an LLM live market-implied probabilities ("what are the odds the Fed holds rates?").
• Journalism: cite current prediction-market odds with a link to the market.
• Weather trading and forecasting research: compare today's temperature bucket prices on Kalshi and Polymarket with the high observed so far at the exact settlement station, or backtest forecasts against settled markets.

How to use it

1. Click Try for free or Start.
2. Optionally enter a Search keyword (fed rate, bitcoin, election) and choose Venues, Category and Market status.
3. Turn on Include price history if you need time series.
4. Click Start, then download the data or connect it to Make, Zapier, n8n, Google Sheets, webhooks or the API.

Input

Field	Type	Description
mode	enum	markets (default) or weather (Kalshi + Polymarket weather markets with station data).
searchQuery	string	Keyword search, e.g. fed rate, bitcoin. Empty = top markets. Ignored in weather mode.
venues	array	Any of polymarket, kalshi, manifold. Default: all.
category	enum	all, politics, elections, economics, finance, crypto, sports, tech, ai, science, world, culture, weather, health, companies. Mapped to each venue's own tags/categories.
status	enum	open (default), closed (ended/resolved), all.
sortBy	enum	volume24h (default), volume, liquidity, newest, closingSoon.
maxItems	integer	Max markets in total, split evenly across venues. Default 30.
maxMarketsPerEvent	integer	Keep the N most likely markets of each multi-outcome event. Default 5, 0 = all.
includePriceHistory	boolean	Add priceHistory to each market (charged separately). Default false.
priceHistoryRange	enum	1d, 1w (default), 1m, 3m, 1y, max.
weatherCity	string	Weather mode: optional city filter, e.g. NYC, Chicago, London.
weatherType	enum	Weather mode: temperature (default), precipitation (rain and snow) or all.
marketUrls	array	Specific markets instead of search: Polymarket event URLs, Kalshi URLs or tickers (KXFEDDECISION-26OCT), Manifold URLs. In weather mode these markets get station data too.
customTag	string	Advanced: a venue-native tag/category/topic slug that overrides category.

Example input:

{
    "searchQuery": "fed rate",
    "venues": ["polymarket", "kalshi", "manifold"],
    "status": "open",
    "maxItems": 30,
    "maxMarketsPerEvent": 3,
    "includePriceHistory": true,
    "priceHistoryRange": "1w"
}

Weather mode example (resolved NYC temperature markets with the observed high/low):

{
    "mode": "weather",
    "weatherCity": "NYC",
    "weatherType": "temperature",
    "status": "closed",
    "maxItems": 20,
    "maxMarketsPerEvent": 0
}

Output

Each dataset item is one market (type: "market") or a free status row (type: "error"). This is real output (shortened):

{
    "type": "market",
    "venue": "kalshi",
    "marketId": "KXFEDDECISION-26OCT-H0",
    "eventId": "KXFEDDECISION-26OCT",
    "eventTitle": "Fed decision in Oct 2026?",
    "url": "https://kalshi.com/markets/kxfeddecision",
    "question": "Will the Federal Reserve Hike rates by 0bps at their October 2026 meeting?",
    "subtitle": "Fed maintains rate",
    "category": "Economics",
    "status": "open",
    "outcomes": [{ "name": "Yes", "price": 0.83 }, { "name": "No", "price": 0.17 }],
    "yesPrice": 0.83,
    "noPrice": 0.17,
    "lastPrice": 0.83,
    "bestBid": 0.83,
    "bestAsk": 0.84,
    "volume": 3698385.15,
    "volume24h": 198361.38,
    "liquidity": null,
    "openInterest": 1383991.68,
    "currency": "USD",
    "openTime": "2025-09-29T14:00:00.000Z",
    "closeTime": "2026-10-28T17:59:00.000Z",
    "resolution": null,
    "priceHistoryRange": "1d",
    "priceHistoryGranularity": "hour",
    "priceHistory": [
        { "time": "2026-10-06T20:00:00.000Z", "price": 0.83 },
        { "time": "2026-10-06T21:00:00.000Z", "price": 0.84 }
    ],
    "extra": { "seriesTicker": "KXFEDDECISION", "rawStatus": "active", "volumeUnit": "contracts ($1 notional each)" },
    "scrapedAt": "2026-10-07T19:21:15.621Z"
}

{
    "type": "market",
    "venue": "polymarket",
    "marketId": "2589812",
    "eventId": "606422",
    "eventTitle": "Fed Decision in October?",
    "url": "https://polymarket.com/event/fed-decision-in-october-20260617190323537/will-there-be-no-change-in-fed-interest-rates-after-the-october-2026-meeting-20260617190324031",
    "question": "Will there be no change in Fed interest rates after the October 2026 meeting?",
    "subtitle": "No change",
    "category": "Fed",
    "status": "open",
    "outcomes": [{ "name": "Yes", "price": 0.835 }, { "name": "No", "price": 0.165 }],
    "yesPrice": 0.835,
    "noPrice": 0.165,
    "volume": 7807449.95,
    "volume24h": 99830.02,
    "liquidity": 790728.13,
    "currency": "USDC",
    "closeTime": "2026-10-29T03:59:00.000Z",
    "extra": { "conditionId": "0xdf9bf27e…", "clobTokenIds": ["111061902544…", "420387758661…"], "eventVolume": 28414063.58 }
}

{
    "type": "market",
    "venue": "manifold",
    "marketId": "hII6EUzNlU",
    "question": "Fed decision in October?",
    "status": "open",
    "outcomes": [
        { "name": "No change", "price": 0.8528 },
        { "name": "Increase (25 bps or more)", "price": 0.14 },
        { "name": "Decrease (25 bps or more)", "price": 0.0072 }
    ],
    "yesPrice": null,
    "currency": "MANA",
    "extra": { "outcomeType": "MULTIPLE_CHOICE" }
}

Weather mode adds station and observation fields to the same market row. Real output (shortened):

{
    "type": "market",
    "venue": "kalshi",
    "marketId": "KXHIGHNY-26SEP27-T65",
    "eventTitle": "Highest temperature in New York City on Sep 27, 2026?",
    "subtitle": "64° or below",
    "status": "resolved",
    "yesPrice": 1,
    "resolution": "Yes",
    "weatherVariable": "highTemperature",
    "targetDate": "2026-09-27",
    "unit": "F",
    "bucketLow": null,
    "bucketHigh": 64,
    "settlementSource": "The Weather Company",
    "stationId": "KNYC",
    "cliStation": "NYC",
    "stationName": "NEW YORK CITY",
    "stationLatitude": 40.779,
    "stationLongitude": -73.96925,
    "stationElevationM": 27,
    "stationTimeZone": "America/New_York",
    "observedValue": 63,
    "observedHigh": 63,
    "observedLow": 58,
    "observedPrecipIn": 1.58,
    "observedSource": "NWS Daily Climate Report (CLI)",
    "observedStatus": "final",
    "observedInBucket": true,
    "settledValue": 63
}

{
    "type": "market",
    "venue": "polymarket",
    "eventTitle": "Highest temperature in NYC on October 7?",
    "subtitle": "66-67°F",
    "yesPrice": 1,
    "resolution": "Yes",
    "weatherVariable": "highTemperature",
    "targetDate": "2026-10-07",
    "unit": "F",
    "bucketLow": 66,
    "bucketHigh": 67,
    "settlementSource": "NOAA",
    "resolutionSourceUrl": "https://www.weather.gov/wrh/timeseries?site=klga",
    "stationId": "KLGA",
    "stationName": "NEW YORK/LA GUARDIA",
    "stationTimeZone": "America/New_York",
    "observedValue": 67,
    "observedHigh": 67,
    "observedLow": 52,
    "observationCount": 24,
    "observedSource": "METAR/ASOS observations",
    "observedStatus": "complete",
    "observedInBucket": true
}

Field notes

• Prices and probabilities are on a 0–1 scale (0.83 = 83%). For Kalshi, yesPrice is the last trade if it lies inside the current bid/ask spread, otherwise the bid/ask midpoint. Resolved markets show the payout (1 or 0).

• volume units differ by venue: USDC on Polymarket, contracts on Kalshi ($1 notional each), and MANA (play money) on Manifold. See currency.

• Multiple-choice Manifold markets have yesPrice: null; their answers are in outcomes.

• priceHistory is the YES (or first outcome) price over time. On Manifold it is rebuilt from public bets and is available for binary markets only.

• Weather mode: Kalshi's US temperature rules name The Weather Company as the source for a CLIxxx NWS climate site. In our check, Kalshi's settled values matched the official NWS daily climate report (returned here as final) in 200 of 200 markets. METAR-derived values (complete / in-progress) come from hourly and special reports and can be 1° lower than the official daily maximum, because the official value also uses readings between reports. Polymarket's NOAA time-series markets resolve on these same METAR reports.

The dataset has ready-made views in Apify Console: Overview, Outcomes & events, Price history, Weather & stations and Errors & status.

Pricing

Pay per event: you only pay for data you receive. Platform usage is included.

Event	Price	When it is charged
market	$0.001 per market ($1 / 1,000)	Each market saved to the dataset (weather mode: station and observation fields included at no extra cost)
price-history	$0.002 per market history	Only when Include price history is on and history data was returned
Errors / empty results	Free	Venue failures, no matches, invalid URLs

Job	Cost
Default run (30 top markets)	$0.03
100 markets with 1-week price history	100 × ($0.001 + $0.002) = $0.30
1,000 resolved markets (no history)	$1.00
200 weather markets with station data and observations	$0.20

Tip: set Max cost per run in the run options. The Actor stops cleanly at your budget, and if your budget covers a market but not its history, the market is delivered without history (priceHistoryNote explains why).

Use with AI agents (MCP)

This Actor works well as a tool for AI agents through the Apify MCP server. Add it with https://mcp.apify.com?tools=openharvest/prediction-markets-scraper, or let the agent find it with search-actors ("polymarket kalshi odds").

Example call-actor input an agent can use to answer "What are the current odds that the Fed holds rates?":

{
    "searchQuery": "fed decision",
    "venues": ["polymarket", "kalshi"],
    "maxItems": 6,
    "maxMarketsPerEvent": 3
}

The agent then reads question, subtitle, yesPrice, volume24h and url from the dataset. Runs with this kind of input finish in a few seconds and cost under $0.01.

For "What is the market saying about today's high in Chicago, and how hot has it been so far at the settlement station?":

{
    "mode": "weather",
    "weatherCity": "Chicago",
    "maxItems": 12,
    "maxMarketsPerEvent": 0
}

The agent reads subtitle, yesPrice, stationId, observedValue, observedStatus and observedInBucket.

FAQ

Do I need API keys or accounts? No. All three venues publish their market data through public APIs.

Why does maxItems give me fewer markets than I asked for? The limit is split evenly across venues. If a venue has fewer matching markets (e.g. a niche search on Kalshi), you get fewer items, and you are not charged for markets that don't exist. A free NO_RESULTS row tells you which venue had nothing.

What does "Max markets per event" do? Events such as "Democratic nominee 2028" contain dozens of markets, one per candidate. By default only the 5 most likely (highest YES price) are kept, so one event doesn't use up your whole limit. Set it to 0 to get every market.

How far back does price history go? max returns the full history at daily resolution. Ranges up to one month are hourly. For closed markets the window ends at the close time.

Can I scrape specific markets? Yes. Put Polymarket event URLs, Kalshi URLs or tickers (KXFEDDECISION-26OCT, KXFEDDECISION-26OCT-H0) or Manifold market URLs into Specific market URLs.

Does it include order books or trades? Not yet. It returns best bid/ask, last price and the price history. Open an issue if you need full order books.

Which weather markets get observed values? Daily high and low temperature markets on both venues (US stations through the NWS daily climate report or METAR, international airports through METAR, Hong Kong for the current day), daily "will it rain" markets at US stations, and Kalshi monthly rain/snow markets whose rules name a CLIxxx site. Markets that settle on a city-area total, on AccuWeather/Synoptic coordinates (Kalshi hourly markets) or on non-station sources such as hurricanes or earthquakes are still returned with their odds, with observedStatus: "unavailable" and a weatherNote explaining why.

How fast do observations update? METAR reports arrive about every hour (more often in bad weather) and the official NWS daily report is published early the next morning. Schedule the Actor every 15–60 minutes for intraday tracking.

Is Kalshi search official? Kalshi market data comes from the official Trade API v2. For keyword search and ranking, the Actor uses Kalshi's public search endpoint (the one kalshi.com uses). If that endpoint is unavailable, it falls back to paging the official /events endpoint and filters locally.

Is it legal to scrape prediction markets?

This Actor only reads public market data that the venues publish through their own public APIs, plus public weather observations (NWS climate reports and METAR data archived by the Iowa Environmental Mesonet, and Hong Kong Observatory open data). It does not log in, place trades or collect personal data. Polymarket and Kalshi restrict trading in some jurisdictions, but reading public market data is unaffected. Make sure your use of the data complies with each venue's terms of service. This is not financial advice.

Related actors

Other Actors by the same developer:

• CoinGecko, DexScreener & Binance Scraper – Crypto Prices & OHLCV: pair crypto prediction markets with spot prices and candles.
• Google Trends Scraper – Interest Over Time & Related Queries: see whether search interest moves before the odds do.
• Google News Scraper – Full Article Text & Real Publisher URLs: the news behind a price move.
• Hacker News Scraper – Stories, Comments & Who Is Hiring Jobs: tech and AI sentiment next to tech markets.

Support & feedback

Found a bug or a missing field? Open an issue in the Issues tab and we usually reply within a day.