Kalshi historical tick data.
Every book change and every trade, for every series.

Best YES bid and ask with sizes and the spread, one row whenever either changed, derived tick by tick from the rebuilt book: not Kalshi's conflated ticker.

The whole ladder after every book event, in YES terms (the NO bids become the asks, at their sizes), 25 levels a side: wide columns bid_px_1 … ask_sz_25, plus the exchange's own hourly snapshots. Each row is self-contained, so you can start reading mid-day.

Every trade: YES and NO price, size in contracts, the taker's side, block-trade flag, trade id.

Kalshi's own ticker feed, about one update a second per market: last price, YES bid/ask with sizes, cumulative volume and open interest in contracts and dollars.

The CF Benchmarks reference indexes Kalshi's crypto markets settle on (BRTI and the ETH, SOL, XRP, DOGE … real-time indexes), 1-second and 5 Hz, with Kalshi's 60-second averages. One file a day; on L25.

Every trade Kalshi reports from 30 June 2021 to 26 September 2026, for the whole exchange, one file a day. Trades only: Kalshi's historical API has no tick-level order book, so these days carry no l1, l2 or ticker.

From our own capture of Kalshi's market-data websocket, on two independent legs in different regions, merged and deduplicated hourly. Every book delta and every exchange snapshot is kept, so the 25-level book is rebuilt on every change and checked against the exchange's own snapshots; each bundle's README carries the match rate per day.

The book, top of book, trades and ticker start on 2026-09-27 (from 11:00 UTC that day), when our capture began; new days are added daily. Before that, Kalshi publishes trades but no order book, so 2021-06-30 to 2026-09-26 is sold as trades-only history: every trade on the exchange, one file per day.

One book per market in YES terms: Kalshi quotes YES and NO bids, and the NO bids become the asks (price 1 − NO bid, at its size), so nothing on the NO side is lost. 25 levels a side as wide columns (bid_px_1 … ask_sz_25), one row per book event, each row the whole ladder, so you can start reading mid-day. L10 carries the best 10 levels.

On the same plans and at the same prices as Polymarket, bought for Kalshi: Starter ($29/mo: L1, trades and ticker for every series, one series at full depth), L10 ($49/mo: the 10-level book) or L25 ($99/mo: the 25-level book and the CF Benchmarks index feed). Each includes 7 days before you join and every day after; earlier captured days are backfill at the plan's rate. The trades history is $7.50 a month on any Kalshi plan, $316.75 for all of it. Kalshi is its own subscription: hold it alongside a Polymarket or Hyperliquid plan and they add up, monthly or yearly on each. Any free account can pull 2026-10-07, any series, up to 1.0 GB.

Parquet, one file per table per series per day in a table-first tree (one read_parquet reads a whole table), and a README with the schemas, the per-day book-quality numbers and the coverage notes for your exact range. Delivery is a download link, the REST API, or SFTP on L25.