Kalshi Scraper — Markets, Odds, Volume & Order Books

Export live markets from Kalshi, the CFTC-regulated prediction exchange, as clean spreadsheet rows: the question, the price, implied probability, spread, 24-hour volume, open interest, category and close date.

No API key. No account. No wallet. Set a filter, press start, export CSV or JSON.

---

Why this one returns markets you can actually use

Kalshi's public market list is dominated by auto-generated multi-game combo contracts that never trade. Measured on 31 July 2026: paging the flat market endpoint returned 3,000 markets with zero 24-hour volume, no order books and no category.

This Actor walks Kalshi's events instead, which is where the real exchange lives. Same measurement, same day: 6,755 markets, 943 with live 24-hour volume, 4,219 with open interest — every one carrying a real category.

Kalshi's API also offers no sort parameter, so this Actor collects a scan window and ranks it before returning. Ask for the 100 most-traded markets and you get 100 traded markets, not the first 100 the API happened to emit.

---

What you get

Column	Example
title	Will SPD win the next German election?
category	Elections
midPrice	0.095
impliedProbability	9.5
yesBid / yesAsk / spread	0.09 / 0.10 / 0.01
volume24h / volume	8418
openInterest	31204
liquidity	4820.50
closeTime	2030-03-25T14:00:00Z
rulesPrimary	full settlement rules text
url	direct link to the market

Prices are always normalised to dollars (0–1), so a row never mixes cents and dollars. Enable Include order book depth and each market also carries live Yes/No bid ladders, best price first.

---

Typical uses

• Track a topic. Schedule hourly on searchQuery: "fed" and build a probability series.
• Screen for what's live. Sort by open interest — the better liveness signal on Kalshi, since many real markets hold open positions without trading today.
• Compare venues. Pair with a Polymarket export to line the two exchanges up.
• Backtest. Set status to Settled and pull resolved markets with their outcomes.
• Watchlist. Pass exact tickers and get just those, every run.

---

Example input

{
  "mode": "markets",
  "status": "open",
  "sortBy": "volume24h",
  "category": "Economics",
  "minOpenInterest": 100,
  "maxItems": 200
}

Watchlist — only these contracts, filters ignored:

{
  "mode": "markets",
  "tickers": ["KXGERELECTION-29MAR25-SPD"],
  "includeOrderBook": true,
  "orderBookDepth": 10
}

Grouped view — one row per event with every contract nested inside:

{ "mode": "events", "status": "open", "category": "Politics", "maxItems": 50 }

---

Settings that matter

• sortBy — the whole point. volume24h for what's trading now, openInterest for what people actually hold, closeTime for what resolves soonest.
• minOpenInterest — the single most useful filter for cutting dormant contracts.
• category — Politics, Elections, Economics, Financials, Sports, Companies, Climate and Weather, Entertainment, Health, Science and Technology, World, Social, Transportation.
• maxItems — your cost ceiling. Set it deliberately.
• useFlatMarketList — off by default, and it should stay off. It enumerates every contract including the dead auto-generated ones. Only turn it on if that is genuinely what you want.

---

Pricing

Pay per event — you pay for rows returned, not for runtime.

Event	Price
Actor start	$0.002 per run
Market or event row	$0.0015 each
Order book (per market)	$0.0008 each

1,000 markets ≈ $1.50. Order books are only charged when you switch them on.

---

Reliability

Plain HTTP against Kalshi's documented public API — no headless browser, no bot-detection workarounds. Runs are fast and cheap, and the Actor doesn't break when a web page changes.

Every request retries with exponential backoff, and a failed page degrades the run instead of ending it. If your filters match nothing, the log tells you which filter to loosen rather than handing back a silent empty dataset.

---

Notes

• Data comes from Kalshi's public trade API and is provided as-is for research and analysis. It is not financial advice and carries no warranty of accuracy or timeliness.
• A row is a snapshot at scrapedAt, not a live feed. Schedule the Actor if you need a series.
• Many Kalshi markets legitimately have zero 24-hour volume. That is the market, not a bug — use minOpenInterest to filter for the ones with real positions behind them.
• Kalshi restricts trading by jurisdiction. Reading public market data is a separate matter from trading; check your local rules before acting on anything here.

---

Questions or a bug?

Open an issue on the Issues tab with your input JSON and the run ID — that is usually enough to reproduce it straight away.