Kalshi Brings Institutional Trading Capabilities with Multicast Data Feed

Kalshi, a prediction market platform, has become the first of its kind to publish its live order book over a dedicated fiber-optic network using multicast distribution. This move is significant because it brings institutional-grade trading capabilities to prediction markets, which until now have been limited by their data plumbing.

Until this week, any firm wanting a continuous view of Kalshi's order book faced a structural problem baked into the platform's data architecture. The company offered price data primarily through REST APIs, a request-response model that has inherent limitations at institutional scale. This architecture breaks in two predictable ways when institutions push it hard: rate limits cap how often systems can poll, and polling produces book drift, causing mispricing under volatile conditions.

The consequence is that systematic market-making on prediction markets has been harder than on regulated equity or futures venues. The gap was not incidental, it shaped who could operate profitably in these markets and at what scale. The architecture that DoubleZero Edge applies to Kalshi's data is called multicast, which enables one-to-many efficiency at scale.

Kalshi's feed on DoubleZero Edge launches with two tiers of data: Level 1, showing the best available bid and ask prices and a record of completed transactions; and Level 2, exposing every resting buy and sell order at every price level in the market, the full ladder of outstanding liquidity.

Firms seeking access to the Kalshi feed can subscribe at doublezero.xyz. The data is published by Kalshi Research, a dedicated unit of the exchange. At launch, sports and crypto perpetual futures account for the majority of Kalshi's weekly notional trading volume.