Kalshi has a public trading API: REST endpoints for markets, order books, placing and cancelling orders and reading your portfolio, plus websocket streams for live prices and fills. You authenticate with API keys generated in your Kalshi account, test against a demo environment with fake money, and pay the same fees as in the app, which means free resting orders on most markets. It's the backbone of the bots and dashboards you see discussed on r/Kalshi.
What you can do with it
- Market data: list markets and events, read order books, trades and candlesticks.
- Trading: place limit and market orders, cancel, amend, batch.
- Portfolio: positions, balance, fills, settlements.
- Streams: websocket subscriptions for order-book updates, trades and your own fills.
Getting started
- Open a Kalshi account and complete verification.
- Generate an API key in account settings; store the private key safely.
- Point your code at the demo environment first. It mirrors the production API with pretend funds.
- Read the rate-limit tier you're on; higher limits are granted on request for serious volume.
- Go live. Fees apply exactly as in the app: taker 7% × P × (1 − P), makers free on most series (schedule).
What people build
- Market-making bots that rest orders on both sides and earn the spread, fee-free.
- Cross-venue monitors comparing Kalshi and Polymarket prices (see arbitrage).
- Dashboards and alerts for price moves on economics releases.
- Research: pulling full price histories for calibration studies.
Rules
One account, your own money, no insider information, no manipulation. Kalshi surveils trading and has fined and suspended users in 2026. The API doesn't change any of that; it just makes you faster.
Kalshi's API documentation is the authoritative source for endpoints, authentication and limits; this page describes the shape, not the spec.