Kalshi Weather Markets — live prices & settled temperatures

Pull every Kalshi daily high / daily low temperature market (KXHIGH*, KXLOW*) as clean JSON: per-bracket bid/ask, last price, volume, open interest, and — for settled markets — the official settlement temperature. Optionally attach recent trades and the full orderbook.

• Auto-discovers every series Kalshi lists (100+ today: 20+ US cities plus London, Paris, Tokyo, Toronto, Sydney, Dubai, Mexico City…). New cities show up without an actor update.
• Reads Kalshi's official public market-data API — no login, no API key, no scraping of HTML, nothing to break when the website changes.
• Built for reliability: rate-limited, retried with backoff, partial failures never kill the run.
• Correct settlement stations: Chicago = Midway (KMDW), Dallas = DFW, Houston = Hobby (KHOU), NYC = Central Park (KNYC). Getting these wrong costs real money.

vs other Kalshi scrapers

The other Kalshi actors in the store (kalshi-prediction-markets-tracker, kalshi-prediction-market-scraper, kalshi-scraper — 24–29 monthly users each) are general market trackers: sports, elections, economics, weather all mixed together, with no series discovery and no settlement-station mapping. None is built specifically for the 100+ KXHIGH/KXLOW weather markets, so none guards against the Chicago-is-Midway-not-O'Hare class of error this actor is built around.

Input

field	default	notes
cities	["NYC"]	Names, aliases or ICAO codes: NYC, Chicago, London, KLAX, EGLL… or ["all"]
seriesType	both	KXHIGH (daily highs), KXLOW (daily lows), both
mode	live	live = tradeable now, settled = finalized with settlement temp, both
lookbackDays	30	settled-history window, 1–365
includeTrades	false	up to 1,000 recent fills per market (+1 call per market)
includeOrderbook	false	current YES/NO bid levels (+1 call per market)
maxRecords	0	stop after N records (0 = unlimited)

Output — one record per market

{
  "ticker": "KXHIGHNY-26SEP16-B81.5",
  "event_ticker": "KXHIGHNY-26SEP16",
  "city": "New York",
  "station_id": "KNYC",
  "series_ticker": "KXHIGHNY",
  "series_type": "KXHIGH",
  "title": "Will the maximum temperature be 81-82° on Sep 16, 2026?",
  "bracket_label": "81° to 82°",
  "strike_type": "between",
  "floor_strike": 81,
  "cap_strike": 82,
  "status": "finalized",
  "result": "no",
  "settlement_temp_f": 77.0,
  "settlement_time": "2026-09-17T12:01:51Z",
  "open_time": "2026-09-15T14:00:00Z",
  "close_time": "2026-09-17T05:00:00Z",
  "yes_bid": 0.0, "yes_ask": 1.0, "no_bid": 0.0, "no_ask": 1.0,
  "last_price": 0.01, "previous_price": 0.02,
  "volume": 4537.22, "volume_24h": 120.0, "open_interest": 4163.05, "liquidity": 0.0,
  "scraped_at": "2026-09-18T08:50:02Z"
}

Prices are USD per contract (0.00–1.00). Quantities can be fractional. result and settlement_temp_f are null while a market is live. When includeTrades / includeOrderbook are on, records carry trades: [{trade_id, created_time, yes_price, no_price, count, taker_side}] and orderbook: {yes_dollars: [[price, size], …], no_dollars: [[price, size], …]}; the keys are absent when the flags are off, and null if that one fetch failed.

Typical sizes

• Live, one city, both series: ~12 records.
• Settled, one city, 30 days: ~350 records. All US cities, 90 days: ~20,000.
• Kalshi lists many international series that only run part of the year; those are reported as "no markets right now" rather than as errors.

Use cases

Backtesting weather-market strategies, calibrating forecasts against settlement temperatures, monitoring bracket prices vs. NWS/METAR observations, feeding dashboards and AI agents.

Pricing

Pay per market record. No subscription, no minimum. See the pricing tab.

Notes

Market data only — this actor never logs in, never places orders, and holds no credentials. Kalshi's API is the source of truth; if a field disappears upstream, the corresponding output field is null rather than the run failing.