Keynote: DFlow (Nitesh Nath)

DFlow has just unveiled one of the most ambitious bridges between traditional finance and crypto: a Prediction Markets API that tokenizes positions from Kalshi, one of the world's most liquid regulated prediction market exchanges, making them accessible to Solana traders for the first time.

Summary

DFlow, a leading low-latency trading infrastructure provider on Solana, has announced the release of their Prediction Markets API powered by Kalshi at Breakpoint 2025. The announcement comes after a breakout year for the company, during which they processed over $33 billion in trading volume and paid out more than $34 million in revenue to application developers since April 2025.

The new API serves as connective tissue between Kalshi's regulated off-chain prediction markets and Solana's highly active trading community. By tokenizing Kalshi positions as SPL tokens, DFlow is enabling a new class of DeFi composability that could spark innovation in social trading, yield products, and applications that haven't even been imagined yet.

What makes this particularly exciting is DFlow's deliberate choice to tokenize these positions rather than simply providing API access. This decision opens the door for developers to build lending protocols around prediction market positions, create novel liquidity mechanisms, and develop entirely new financial products—all while maintaining a connection to one of the most legitimate prediction market exchanges in existence.

The company already has over 100 integrations lined up for the prediction markets API, with many expected to go live in the coming quarter and some announcements happening the same day as the keynote.

Key Points:

DFlow's Explosive 2025 Growth

DFlow's DEX aggregator business experienced remarkable growth throughout 2025. Starting with just four application integrations in June, the platform expanded to 26 integrations by September and surged to over 75 applications by November. This growth was accompanied by massive volume increases—from $4 billion in cumulative volume in June to $33 billion by November.

The revenue numbers are equally impressive. Application developers using DFlow's infrastructure earned $2 million in cumulative revenue by June, which ballooned to $34 million by November. This demonstrates not just adoption, but meaningful economic value flowing to builders in the ecosystem.

Just-In-Time Routing: Solana's Microstructure Innovation

DFlow released what Nitesh Nath called "Solana's most exciting microstructure improvement of 2025"—just-in-time routing. This zero-to-one mechanism fundamentally changed how transactions work on the network by making previously frozen transactions dynamic at the time of on-chain processing.

The practical impact is significant: users experience dramatically better execution, reduced price slippage, and more value returned to their wallets. This innovation also enabled applications to capture additional revenue from improved execution. The feature was particularly impactful when combined with DFlow's new prop AMM integration backend, allowing proprietary automated market makers to integrate smoothly and access features unavailable elsewhere.

How Tokenized Prediction Markets Work

The technical implementation of tokenized prediction markets relies on what DFlow calls a "concurrent liquidity program." This program essentially glues off-chain liquidity from Kalshi to the Solana network, allowing users to receive SPL tokens representing their prediction market positions.

The process works through a multi-transaction flow. When opening a position, the first transaction escrows user funds and increases their position. The second transaction facilitates the actual trade and mints tokens representing the user's prediction market stake. A similar multi-step process handles redemption—when a prediction resolves correctly, users burn their position tokens and receive stablecoins in return.

The Strategic Bet on Tokenization and Composability

DFlow's decision to tokenize Kalshi positions rather than simply providing direct API access was a deliberate choice driven by the team's belief in DeFi's composable nature. As Nath explained, they didn't have a complete picture of why tokenization was important but understood the "serendipitous nature of tokens and DeFi."

This bet on serendipity is essentially a bet on the creativity of Solana builders. The team expects to see prediction tokens used in bar lending protocols, new liquidity formation mechanisms, and applications that haven't been conceived yet. The company drew parallels to past innovations like pump.fun and prop AMMs that emerged unexpectedly to reshape the ecosystem.

100+ Integrations Ready to Launch

The appetite for tokenized prediction markets is already enormous. DFlow has lined up over 100 integrations, with most expected to go live in Q1 2026 and many launching within weeks of the announcement. Builders have expressed interest across multiple categories including social trading applications, niche market-focused platforms, data vendor integrations, group trading via Telegram, and yield products built on prediction market positions.

Facts + Figures

• DFlow has processed over $33 billion in trading volume since April 2025
• More than $34 million has been paid out in application revenue to date
• The platform grew from 4 integrations in June 2025 to over 75 by November 2025
• Over 1 million users globally have been serviced by DFlow
• The DEX aggregator covers 100% of liquidity on the Solana network
• Over 100 integrations are lined up for the new Prediction Markets API
• Application revenue grew from $2 million in June to $16 million by September to $34 million by November
• Cumulative volume grew from $4 billion in June to $10 billion in September to $33 billion by November
• The Prediction Markets API was released just days before the Breakpoint presentation
• Just-in-time routing was released in September 2025

Top Quotes

• "This year has been pretty phenomenal for our growth. We have facilitated over $33 billion in trading volume since April 2025."

• "Just-in-time routing was a zero-to-one mechanism that was released on the network."

• "Solana users will begin to receive market access to one of the most liquid prediction markets exchanges in the world. In exchange, Kalshi receives access to some of the most automated, active degens that the world has ever seen."

• "Tokenization was a choice. In building this prediction markets API, we did not need to tokenize. But we felt it was really important to do."

• "We understood the serendipitous nature of tokens and DeFi. The fact that you can compose liquidity, compose tokens, and have novel mechanisms designed by anyone was just phenomenal."

• "So we took a bet on serendipity, and this is also a bet on the imagination of builders on Solana."

• "We have no idea what is coming next, but we're very excited for it."

Questions Answered

What is DFlow and what does it do?

DFlow builds low-latency trading infrastructure on Solana. Concretely, they operate a DEX aggregator that covers 100% of liquidity on the network. The system constructs a large graph of prices and solves complex optimization problems to determine the best trading routes from the token a user wants to sell to the token they want to receive. Their infrastructure powers swap functionality across more than 75 applications on Solana, enabling users to access the best available prices across all decentralized exchanges on the network.

How does the tokenized prediction market system work technically?

The system uses what DFlow calls a "concurrent liquidity program" that connects off-chain Kalshi liquidity to Solana. When a user opens a prediction market position, the first transaction escrows their funds and increases their position. The second transaction completes the fill and mints SPL tokens representing the user's stake. For redemption, users burn their position tokens, and if their prediction was correct, they receive stablecoins. This multi-transaction flow ensures that on-chain tokens always accurately represent off-chain positions.

Why did DFlow choose to tokenize prediction market positions instead of just providing API access?

DFlow tokenized positions because they believe in the composable nature of DeFi and wanted to enable unexpected innovation. While tokenization wasn't technically necessary, the team recognized that SPL tokens can be integrated into lending protocols, used in novel liquidity mechanisms, and composed in ways that pure API access cannot enable. They explicitly described this as "a bet on serendipity" and on the creative potential of Solana developers to build applications that the DFlow team cannot yet imagine.

What is just-in-time routing and why is it significant?

Just-in-time routing is a mechanism DFlow released in September 2025 that makes previously frozen Solana transactions dynamic at the moment of on-chain processing. Before this innovation, transaction parameters were set when submitted and couldn't adapt to market conditions. With just-in-time routing, transactions can adjust in real-time, resulting in significantly better execution for users, reduced price slippage, and additional revenue opportunities for applications. DFlow's founder described it as Solana's most exciting microstructure improvement of 2025.

What kind of applications are developers building with the Prediction Markets API?

Developers have expressed interest across diverse categories. Social trading applications that let users follow and copy prediction market strategies are common. Niche market-focused platforms that specialize in specific types of predictions are in development. Data vendors want to incorporate prediction market information. Group trading applications, particularly via Telegram, are being built to enable communities to trade together. Yield products that generate returns from prediction market positions round out the use cases builders are exploring.

How does DFlow benefit application developers financially?

Applications that integrate DFlow's APIs receive revenue from the trading activity they facilitate. Since April 2025, DFlow has paid out over $34 million in cumulative application revenue. This revenue comes from improved execution and the economic activity enabled by DFlow's infrastructure. The growth in payouts—from $2 million in June to $34 million by November—demonstrates how rapidly this revenue stream can scale as integration adoption increases.