Kalshi backtesting data, down to the order book.

Recorded Kalshi yes/no ladders, up to 100 levels per side. Test fees, slippage, partial fills, and depth-aware execution on stored book observations.

Get a free API keyStart free
1.5B
yes/no books captured
2.6M+
markets
100
depth levels
REST
capture method
Backtest Lab

Find out whether an idea has an edge, no pipeline needed. The Backtest Lab runs the whole test in the browser, on the real recorded book.

Paper trading

A backtest is a claim about the past. Paper trading tests it on markets still to come: virtual cash, real prices, a record you cannot fake.

Most feeds stop at the last price.

A single number. It says nothing about the size waiting to fill.

We capture every level, both sides.

The full bid and ask ladder, with the size resting at every price.

Recorded at every change, forever.

Every frame stored, so a backtest fills against the liquidity that was really there.

This is where DepthFeed comes in.

The book, not just the last price.

The full book, every level, both sides

  • Every level, with size
  • Both sides of the book
  • Captured at every change

Polled at full depth, continuously

  • Public REST orderbook
  • Continuous full-depth polls
  • Finer than hourly archives

One REST API for latest and history

  • Latest and past snapshots
  • JSON, epoch-millis stamps
  • Keyset pagination

Underlying prices, time-aligned

  • Binance spot and futures
  • Chainlink settlement marks
  • Joined on epoch millis

The order book as it was, on every prediction market

DepthFeedStudio

Backtest on the full order book.

Replay the book as it was.

Size the slippage first.

Then pull it over the API.

View Studio

Backtests

Liquidity

Slippage

Market catalogue

For

  • Quant researchers
  • Systematic traders
  • Market makers
  • Prop desks
  • Data scientists
  • Risk teams
  • Academic researchers
  • Strategy builders
  • Trading bots
  • Anyone testing an edge

Full order-book depth, captured continuously. One schema across every venue we cover.

Kalshi

Start backtesting on real depth.

Start free