I’ve been watching prediction markets mature for a while now, and every so often something lands that feels less like incremental progress and more like a structural shift. This week it arrived in the form of a quiet but important partnership: Kalshi has started feeding its live order book directly into DoubleZero Edge. That means Level 1 and Level 2 data for the platform’s most active sports contracts and crypto perpetual futures now travel over a dedicated fiber network instead of the usual public-internet scramble. For anyone who has ever tried to rebuild an order book from raw API responses, the difference is immediately obvious.
Why Live Market Data Distribution Finally Matters in Prediction Markets
Prediction markets spent years operating in a kind of data wilderness. Traders and quant desks pulled snapshots, pieced together partial books, and hoped the latency wouldn’t kill the edge. Traditional exchanges solved this problem decades ago with multicast feeds and private networks. Crypto and event markets largely skipped that infrastructure step. DoubleZero’s approach tries to close the gap by treating market data the same way established venues treat it: as a product that needs reliable, sequenced delivery rather than something you scrape and reconstruct yourself.
The rollout covers Top of Book and Trades (what most people call Level 1) plus full Depth of Book (Level 2). Level 1 shows the best bid and offer along with the last trade. Level 2 layers in the resting orders at multiple price levels so you can actually see liquidity rather than just the headline price. That extra visibility matters when you’re sizing positions or running market-making logic. In my experience, the difference between a thin top-of-book view and a genuine depth feed is the difference between guessing and measuring.
How the Feed Actually Works
DoubleZero Edge does not blast individual copies of the data to every subscriber. It uses multicast. The publisher sends the stream once; the network then fans it out simultaneously to every connected participant. That model is familiar to anyone who has worked with equity or futures market data. It reduces the load on the origin and keeps the sequence consistent across receivers. Kalshi Research supplies the source information; DoubleZero handles the transport over its fiber routes instead of relying solely on the open internet.
Before this arrangement, firms often had to poll Kalshi’s APIs, stitch responses together, and maintain their own book-building logic. Edge packages the data into a ready subscription product. You still need to connect and meet the network requirements, but a large piece of the internal plumbing disappears. Neither side has published exact latency numbers or the list of initial customers, which leaves some questions open. Still, the architecture itself is a clear step toward the kind of infrastructure traditional markets take for granted.
The Commercial Twist That Caught My Attention
Here is the part that feels unusual. For the first year Kalshi is waiving its normal share of Edge subscription revenue. Data publishers on the network typically receive a percentage after protocol fees. By forgoing that cut, Kalshi effectively lets the initial price reflect pure network delivery costs rather than an added data-licensing layer. Subscribers still pay DoubleZero’s network fee and must satisfy connection requirements, so free access is not on the table. But the absence of the publisher take for twelve months is a deliberate incentive to seed adoption.
Once the waiver ends, the standard revenue-share model can resume. The announcement left the exact percentage and any future price impact unspecified. I’ve seen similar introductory periods in other market-data products; they often succeed when the underlying demand is already present. And demand does appear to be present. Combined July volume across major prediction platforms hit roughly 50.59 billion in taker notional, with Kalshi accounting for the large majority of that figure. That kind of activity creates a natural audience for cleaner, lower-latency data.
What Level 1 and Level 2 Actually Deliver
It helps to be precise about the content of the feed. Level 1 gives you the best available bid and ask plus trade prints. That is enough for many display applications and simple strategies. Level 2 expands the picture to multiple price levels on both sides of the book. Market makers and quantitative desks use that depth to estimate liquidity, model impact, and decide whether a given size can be absorbed without excessive slippage. The data arrives in a sequenced, machine-readable format designed for direct integration into pricing and hedging systems.
Historical data is missing from the first release. DoubleZero has said it plans to add historical Kalshi information later, but no timeline or pricing has been shared. For firms that rely on back-testing or longer-term analytics, that gap is noticeable. Real-time depth is the priority for now, and that focus makes sense given the speed at which these markets move.
From Blockchain Packets to Prediction-Market Books
DoubleZero did not start with prediction markets. Earlier this year the network launched a public beta focused on Solana transaction data. Hundreds of validators representing a substantial share of staked supply began publishing raw packets over private fiber. Measured delivery improvements versus conventional routing were modest in absolute terms yet consistent enough to matter at scale. Subscription pricing for that beta sat in a relatively accessible range depending on location and device count. Whether the same structure applies to the Kalshi product remains unclear, but the technical foundation is the same: dedicated fiber, multicast distribution, and an emphasis on reducing the variability that public internet paths introduce.
Austin Federa, one of DoubleZero’s co-founders, has framed the effort as bringing a concept that traditional finance already understands well into newer markets. Data access, he argues, is a core piece of market structure rather than an afterthought. Prediction platforms and perpetual futures venues grew up without the shared private-network layer that equity and futures firms spent years building. The Kalshi feed is an attempt to import that layer rather than reinvent it.
Traditional finance got this concept exactly right: data access is a critical part of market structure.
That statement lands differently when you remember how fragmented crypto and event-market data still are. Many venues still expect participants to handle book reconstruction themselves. Offering a packaged, sequenced feed over fiber is a meaningful departure from that status quo.
Crypto Perpetuals and the Regulated Path
The crypto side of the feed includes perpetual futures that Kalshi introduced after receiving the necessary regulatory clearance. Bitcoin perpetual contracts track the spot price and remain open-ended rather than expiring on a fixed schedule. Additional crypto-linked perps followed. For eligible U.S. participants these products opened a domestic route to derivatives that had previously been concentrated on offshore platforms. Having the order books for those contracts available through the same Edge connection creates a single data path for both sports event markets and regulated crypto derivatives.
Kalshi has operated as a designated contract market for several years. Later modifications expanded the scope of intermediated futures trading. Federal registration does not automatically resolve every state-level question around sports contracts. Some jurisdictions continue to assert local gambling authority over certain event markets. Those legal tensions sit outside the data-feed story, yet they form part of the broader environment in which the markets operate. The data itself, however, is now traveling on infrastructure that looks more like the rest of institutional finance.
Practical Implications for Different Users
Quantitative firms and market makers gain the most immediate benefit. Full book depth in a consistent, low-variability stream reduces the engineering overhead of maintaining accurate local books. Automated pricing and hedging systems can consume the feed more cleanly. Even discretionary traders who rely on depth charts rather than raw APIs stand to see a more stable picture of liquidity.
Smaller participants may notice the improvement less dramatically if they already work through existing interfaces. The subscription still requires a connection to the Edge network, so casual retail users are unlikely to be the primary audience. The product is aimed at firms that treat data as infrastructure rather than a convenience feature.
- Lower variability in delivery times compared with pure public-internet paths
- Sequenced Level 1 and Level 2 data ready for machine consumption
- Coverage of both sports event contracts and crypto perpetual futures
- First-year publisher revenue waiver that keeps the initial price closer to network cost
- Future historical data planned but not yet scheduled
I’ve found that the real test of any new market-data product is not the launch announcement but the quiet months afterward. Do the books stay clean under load? Does the sequence hold when activity spikes? Are the connection requirements practical for the desks that need the data most? Those answers will arrive through use rather than press releases.
The Broader Context of Rising Volumes
Prediction-market activity climbed noticeably through the summer. July notional volume across the main platforms reached a combined figure north of 50 billion, with one venue responsible for the large majority. That growth creates both opportunity and pressure. Higher volume attracts more sophisticated participants who expect institutional-grade data. At the same time it increases the cost of maintaining accurate books if every firm has to rebuild them independently. A shared, multicast feed is one way to absorb that pressure without forcing every participant to reinvent the same infrastructure.
Perhaps the most interesting aspect is how cleanly the Kalshi feed sits inside DoubleZero’s existing model. The network already carried blockchain data. Extending the same distribution layer to a regulated prediction-market order book feels like a natural expansion rather than a forced pivot. Whether other venues follow with their own feeds remains an open question, but the precedent is now set.
What Remains Unclear
Several practical details are still missing. Exact subscription pricing for the Kalshi product has not been disclosed. Measured latency figures relative to direct API access or competing services are likewise absent. The number of early adopters is unknown. Historical data is promised for a later release without a date. These gaps are common at the start of a new data product; they simply mean the full picture will emerge through operation rather than through the initial announcement.
For firms already running strategies on Kalshi markets, the decision reduces to a cost-benefit calculation: is the reduction in engineering overhead and delivery variability worth the network fee? For those still evaluating the markets, the existence of a more structured data path may lower one barrier to entry. Either way, the conversation has moved beyond “can we get the data at all” to “how cleanly and consistently can we receive it.”
Looking Ahead Without Over-Promising
Market infrastructure rarely changes overnight. What looks like a simple feed announcement often turns out to be the visible tip of longer work on routing, sequencing, and commercial models. DoubleZero’s fiber approach and Kalshi’s willingness to open its book under a temporary revenue waiver form one concrete step. Whether it becomes a template for other prediction markets or remains a single-venue solution will depend on adoption and on how well the product performs under real trading conditions.
I’ve noticed that the most durable improvements in market data tend to be the ones that feel boring once they settle in. People stop talking about the feed and simply use it. If the Kalshi data on DoubleZero Edge reaches that stage, the real success will be measured in quieter codebases and more consistent books rather than in headlines. For now the feed is live, the first-year commercial terms are favorable to early subscribers, and the architecture follows principles that traditional markets already trust. That combination is worth watching closely as volumes continue to grow and as more firms treat prediction-market data with the same seriousness they apply to every other asset class.
The practical next question for any desk is straightforward: does the current workflow already deliver clean enough depth, or is the engineering and latency cost of maintaining that depth still higher than necessary? The answer will differ by firm size, strategy, and existing infrastructure. What has changed is that a packaged alternative now exists. In a market that spent years without one, that fact alone shifts the baseline.
One final observation. Prediction markets have always lived at the intersection of information and capital. Better data distribution does not invent new information, but it does reduce the friction between the information that exists and the capital that wants to act on it. When that friction drops, pricing tends to become a little sharper and participation a little broader. Whether the Kalshi–DoubleZero connection produces that effect at scale is an empirical question that only time and trading volume can answer. The infrastructure for testing the hypothesis is now in place.
As volumes keep climbing and more traditional firms look at these markets, the presence of a familiar data-delivery model may matter more than many currently expect. Fiber, multicast, sequenced books—these are not glamorous concepts. They are, however, the quiet machinery that lets sophisticated strategies operate without constant reconstruction work. Bringing that machinery into prediction markets is less a revolution than a long-overdue alignment. And alignment, in market structure, often proves more powerful than novelty.
For anyone building or trading in these markets, the practical takeaway is simple. A new distribution path for live order-book data exists. It covers the contracts that already attract the most activity. The commercial terms for the first year remove one common friction. The rest will be decided by usage, reliability, and whether the promised historical layer eventually arrives. Those are the metrics that matter once the announcements fade.
