Key Points
The core logic of prediction markets is to aggregate dispersed information and opinions through price mechanisms, making financial bets a true reflection of the probability of event outcomes. Because trading incurs costs, participants are more rational in their betting, allowing market prices to embody collective wisdom.
Although early prediction markets proved the effectiveness of price mechanisms, they were constrained by regulatory and technological limitations, preventing large-scale development. The emergence of Polymarket marked the true explosion of prediction markets, adopting a hybrid model of off-chain matching and on-chain settlement, significantly enhancing trading efficiency and experiencing explosive growth during the 2024 U.S. elections.
Currently, the centralized core prediction market representative Kalshi has surpassed a monthly trading volume of $1.9 billion, with a market share of approximately 63.51%; the decentralized core prediction market representative Polymarket has a monthly trading volume of about $1.1 billion. At the same time, the emergence of innovative projects such as Limitless, Gondor, Melee, Polyfactual, and Flipr has accelerated competition in the entire sector.
From the perspective of prediction market categories, sports and crypto have become the main growth engines; political categories remain representative but are constrained by regulations; while entertainment and social events have potential for growth. For prediction markets to achieve sustained growth, they can deeply integrate with rapidly developing on-chain ecosystems such as DeFi, AI, social, aggregation protocols, TradeFi, and Meme to further enhance data density and interaction depth.
Prediction markets have shifted from mere event speculation to price discovery based on information value and quantitative expression of social expectations. However, issues such as regulatory uncertainty, insufficient liquidity, and user speculation still need to be addressed. Only through continuous innovation in event generation mechanisms, leverage structures, and information valuation can prediction markets truly evolve from speculative platforms to pricing layers of social expectations, thereby becoming the underlying infrastructure linking cognitive systems and financial pricing mechanisms.
Table of Contents
I. The Operation Mechanism of Prediction Markets
1.1. The Basic Logic of Prediction Markets
1.2. Blockchain and Prediction Markets
1.3. The Differences of Prediction Markets
II. The Development Process of Prediction Markets
2.1. Early Exploration Stage
2.2. The New Generation After the DeFi Boom
2.3. A New Cycle of Compliance and Commercialization
III. Analysis of Core Prediction Platforms
3.1. Comprehensive Types
3.2. DeFi Types
3.3.AI Types
3.4. Other Types
IV. Data and Market Structure
4.1. Market Overview
4.2. User Analysis
4.3. Major Categories
V. Development Path and Potential Expansion
5.1. Integration with DeFi
5.2.AI Driven
5.3. Social Extension
5.4. Protocol Aggregation
5.5. Bridging TradeFi
5.6. Attention Restructuring
VI. Industry Landscape and Future Positioning
6.1. Challenges and Constraints
6.2. Future and Growth Landscape
Reference Links
I. The Operation Mechanism of Prediction Markets
1.1. The Basic Logic of Prediction Markets
The core of prediction markets lies in utilizing price mechanisms to aggregate dispersed cognition and information. Unlike traditional polls, prediction markets require participants to bet funds on the outcome of an event, with the price serving as a dynamic estimate of the probability of the event occurring. Since betting incurs opportunity costs, participants will only act when they believe they have an information advantage, and this incentive structure encourages more rational judgments, reducing noise from costless opinions, thereby allowing prediction markets to extract clearer signals in complex environments. Furthermore, prediction markets are not only a type of trading mechanism but are also seen as a form of collective decision-making. When participants possess diversity, independence, and decentralized information sources, the price mechanism becomes an effective aggregation tool, integrating the judgments of different individuals into a more accurate collective probability. Therefore, the significance of prediction markets is not limited to fund-driven betting behavior but is more reflected in their ability to achieve efficient aggregation of collective cognition through market mechanisms. Additionally, from a macro perspective, prediction markets play a role as a reference for social decision-making, with price trends reflecting public confidence in policies, elections, economic indicators, and even technological breakthroughs. This market-based prediction is more real-time and continuous compared to traditional research methods, making it particularly suitable for fields with rapidly changing information.
1.2. Blockchain and Prediction Markets
Blockchain provides a publicly transparent and immutable operational foundation for prediction markets, resolving the trust dilemmas of traditional platforms. By using smart contracts to escrow funds and automatically settle results, users do not need to rely on centralized institutions, effectively reducing moral and default risks. This "code is trust" mechanism not only enhances participants' sense of security but also builds a permissionless fair market for global users. Meanwhile, blockchain technology significantly improves the efficiency and scalability of prediction markets. The combination of decentralized oracles and smart contracts enables automated, verifiable event settlements, reducing human intervention and ensuring fair outcomes; the proliferation of cryptocurrencies lowers the barriers for cross-border transactions and small bets, further enhancing liquidity and user experience. Blockchain allows prediction markets to become not only permissionless and openly shared global infrastructure but also capable of large-scale collaboration and information aggregation.
1.3. The Differences of Prediction Markets
The biggest difference between prediction markets and traditional gambling lies in their goals and information value. Gambling odds are set by the house, with participants primarily engaging for entertainment or speculation, resulting in a binary win or lose outcome; whereas prediction markets cover a broader range of topics such as politics, economics, and society, emphasizing the probabilities carried by prices. The market price is not only the result of betting but is also seen as a reflection of collective wisdom, providing references for policy-making and business decisions. Compared to financial derivatives and voting markets, prediction markets also exhibit a unique positioning. Derivatives rely on specific financial assets, primarily serving risk hedging and speculation, while prediction markets can transform almost any objective event into a trading target. As for voting and polling, which follow the principle of "one person, one vote," they are often lagging and struggle to dynamically reflect changes; prediction markets, on the other hand, provide more sensitive probability signals through capital weighting and real-time prices. Prediction markets combine the intuitiveness and fun of gambling with the rational trading mechanisms of financial markets, while also fulfilling a similar information aggregation function as polls.
Source: CoinW Research Institute
II. The Development Process of Prediction Markets
2.1. Early Exploration Stage
The concept of prediction markets has a long history, dating back to betting activities during the papal elections in the 16th century. The modern electronic prediction market's prototype is the Iowa Electronic Markets launched by the University of Iowa in 1988, used to predict presidential elections. In the 1990s, some online prediction platforms emerged, such as InTrade and BetFair, attempting to allow the public to bet on political, entertainment, and other events using real money or game tokens. These platforms validated the effectiveness of collective predictions to some extent but gradually faded out due to legal and operational issues. The real integration of prediction markets with cryptographic technology came with Augur. Augur raised funds through an ICO in 2015 and became one of the earliest decentralized prediction market protocols on Ethereum, officially launching its mainnet in 2018. However, Augur's early practices exposed many issues. At that time, the Ethereum main chain was slow and had high transaction fees, making the costs of creating and trading markets on Augur prohibitive. Ultimately, Augur had very few daily active users, and its REP token's market value significantly declined. In addition to Augur, there were concurrent attempts such as the Gnosis project. Gnosis initially positioned itself as a prediction market platform but later shifted to developing decentralized exchanges and management tools, with its team incubating prediction applications like Omen, which also launched but had limited impact. Prediction markets during this stage failed to break through in underlying performance and product design, failing to attract public participation.
2.2. The New Generation After the DeFi Boom
The rise of DeFi has brought new vitality to prediction markets. After 2019, the maturity of DeFi infrastructure, including stablecoins, decentralized wallets, and cross-chain settlement protocols, provided a composable financial foundation for the operation of prediction markets. This transformed prediction markets from mere speculative tools into platforms capable of efficiently allocating capital and risk, similar to other DeFi protocols. In this context, Polymarket has become a representative of the new generation of prediction markets. It adopts a hybrid architecture of off-chain matching and on-chain settlement, avoiding the high gas costs of Ethereum while maintaining transaction security and verifiability. Users can directly use stablecoins like USDC to place bets, significantly lowering the participation threshold. Meanwhile, DeFi's liquidity pools and yield aggregation mechanisms provide prediction markets with higher capital efficiency and instant settlement capabilities. With the success of Polymarket, more new platforms have chosen to build on high-performance public chains or layer two networks, further DeFi-ifying their mechanisms, such as introducing LP mechanisms. These innovations not only improve user experience and capital utilization but also enhance the capital attractiveness of prediction markets. The heat after 2024 indicates that prediction markets are transitioning from niche applications to open financial infrastructure with macro pricing and risk hedging functions.
2.3. A New Cycle of Compliance and Commercialization
In recent years, the development of prediction markets has entered a new cycle of parallel compliance and commercialization. On one hand, regulatory clarification injects certainty into the industry. Kalshi was the first to obtain CFTC approval, becoming the first legally operating prediction market platform; some regions in Europe and Asia are also exploring regulatory frameworks for "event derivatives." Prediction markets are gradually moving out of the gray area and gaining recognition from the mainstream financial system, paving the way for institutional funds and professional investors to enter. The compliance process will gradually transform prediction markets from marginalized products into regulated financial instruments, which in the long run may complement futures and options as important tools for risk management and sentiment pricing; on the other hand, the wave of commercialization is reshaping the industry landscape. In July 2025, Polymarket acquired the trading platform QCEX, which holds CFTC approval, and partnered with Chainlink. Meanwhile, traditional financial giants are entering the space; on October 7, the parent company of the New York Stock Exchange, ICE, announced plans to invest up to $2 billion in Polymarket, which has a latest valuation of about $9 billion; on October 10, Kalshi completed over $300 million in financing, with its valuation soaring to $5 billion. These significant financings not only bring capital and credibility to prediction markets but also signify that prediction markets are transitioning from niche experiments to mainstream financial infrastructure.
III. Analysis of Core Prediction Platforms
3.1. Comprehensive Types
3.1.1. Centralized Representative Kalshi
Development History
Kalshi was founded in 2018 by former Goldman Sachs quantitative analyst Tarek Mansour and MIT economist Luana Lopes Lara, aiming to achieve the financialization of events within a strict regulatory framework. Both founders have a strong background in derivatives pricing and market structure; Mansour previously served as an executive at the Chicago Mercantile Exchange, while Lara was responsible for emerging market strategies at Morgan Stanley. After nearly three years of preparation, Kalshi received a "Designated Contract Market" (DCM) license from the U.S. Commodity Futures Trading Commission (CFTC) in November 2020, becoming the first platform in the U.S. to legally offer event contract trading at the federal level. After its official launch in 2021, Kalshi was the first to introduce prediction contracts centered around macroeconomic themes such as CPI and non-farm employment.
After obtaining compliance status, Kalshi began to expand into macro and structured markets. In 2022, the platform added various event contracts covering inflation, GDP, energy prices, climate risks, and attracted quantitative funds and institutional investors. However, Kalshi also faced regulatory controversies during this phase. At the end of 2023, the CFTC restricted its "Congressional Control" contracts, citing their "gambling-like nature," which drew widespread market attention and became a catalyst for subsequent regulatory litigation. In 2024, Kalshi shifted its development focus to institutional building and vertical integration. In August 2024, its clearing subsidiary Kalshi Klear obtained a CFTC "Derivatives Clearing Organization" (DCO) license, forming a complete closed loop from trading to settlement, significantly enhancing compliance and risk control capabilities. In October 2024, the U.S. Court of Appeals for the District of Columbia overturned the CFTC's restrictive ruling on election contracts, giving Kalshi the green light to restart its "election market." This ruling marked a significant watershed in the development of event markets, indicating that the compliance of prediction markets was confirmed at the judicial level for the first time. Subsequently, the platform began piloting a broader range of social event predictions, including political, sports, and policy issues.
Since the beginning of this year, Kalshi has entered a phase of rapid capitalization. In June 2025, it completed a $185 million financing round, reaching a valuation of $2 billion. Just four months later, in October, Kalshi completed another over $300 million Series D financing round, co-led by Sequoia Capital and a16z, with participation from institutions like Paradigm and Coinbase Ventures, raising its valuation to $5 billion. Additionally, in January of this year, Donald Trump Jr. joined Kalshi's advisory board, bringing significant traffic and political influence to the election market.
User Data
According to data from The Block, the overall trading volume of prediction markets has significantly increased since 2024, with the monthly total trading volume rising from about $2 billion to nearly $5 billion. Among them, Polymarket dominated from October 2024 to June 2025, with a monthly trading volume ranging from $1.5 billion to $3.5 billion, peaking at about $4 billion at the end of 2024. Meanwhile, Kalshi's trading volume remained below $500 million at the beginning of 2025 but began to rise rapidly from March 2025, surpassing $1.5 billion in August 2025 and further climbing to about $2 billion in September 2025. Currently, Kalshi's monthly trading volume is approximately $1.9 billion.
Source: The Block, https://www.theblock.co/data/decentralized-finance/prediction-markets-and-betting
From the changes in trading volume shares of major prediction markets, Polymarket initially held a long-term dominant position, while Kalshi's market share steadily increased. Over time, Kalshi gradually narrowed the gap with sustained growth in trading volume, reaching 63.51% in September of this year, surpassing Polymarket for the first time. This turning point marks Kalshi's rapid rise in the prediction market field, with its market influence and dominance accelerating.
Source: The Block, https://www.theblock.co/data/decentralized-finance/prediction-markets-and-betting
From Kalshi's weekly nominal trading volume, the market structure is highly concentrated. As of now, the platform's activity is concentrated in three main categories: sports, crypto, and politics. The sports market leads with an absolute trading volume of about $818 million, accounting for nearly 90% of the total, demonstrating the dominant position of sports events in the prediction market; the crypto category follows with $31.84 million, reflecting users' ongoing interest in crypto asset prices and event expectations; the political category ranks third with $20.64 million, although it has declined compared to the peak election cycle, it remains stable and active. These three categories collectively occupy the main liquidity of the platform, forming the core trading structure of Kalshi.
Source: Dune, https://dune.com/datadashboards/prediction-markets
From Kalshi's global traffic source data, the U.S. leads with an absolute traffic share of 78.75%, growing about 120% compared to earlier periods, indicating that the primary active users of the event contract market are still concentrated in North America, especially driven by macro and social events such as the U.S. elections, CPI data, and sports seasons. Following the U.S. are Canada (2.87%), the UK (1.75%), and Germany (1.10%), with respective growth rates of about 52.7%, 48.9%, and 77.4%. Notably, while Turkey's traffic share is only 1.02%, it has grown nearly 892% year-on-year, becoming the fastest-growing overseas market, which is related to the rising interest in alternative speculative assets among the public amid high inflation. Overall, this data reveals that Kalshi remains U.S.-centric, but overseas markets are rapidly penetrating, presenting an early global pattern of North American dominance, European diffusion, and emerging market explosion.
Source: Similarweb, https://secure.similarweb.com
Operational Mechanism
As a designated contract market, Kalshi is regulated by the U.S. Commodity Futures Trading Commission (CFTC), which requires it to fulfill the functions of rule-making, matching, market monitoring, and compliance disclosure as an exchange. Its associated clearing entity, Kalshi Klear, is a registered derivatives clearing organization responsible for post-trade margin management, counterparty risk mitigation, and default handling. This model of combining an exchange with a clearinghouse transforms the originally peer-to-peer betting mechanism into a platform-centered guarantee and net settlement mechanism, significantly reducing counterparty risk.
In terms of contract design and pricing mechanisms, most of Kalshi's products are binary (YES/NO) contracts, with each contract ultimately settling at $1 (if the prediction is correct, the participant receives $1; otherwise, they receive $0). Kalshi supports two types of order placements: market orders and limit orders. In terms of risk control, incentives, and behavioral regulation, Kalshi employs a comprehensive set of rules to ensure market fairness and liquidity. To maintain liquidity, Kalshi provides market maker support and incentive programs (such as fee discounts) and implements monitoring for trading anomalies, matching replay, and activity disclosures; the platform's trading time window may also be affected by system maintenance and other factors.
Business Model
Kalshi's primary revenue source is transaction fees based on expected returns. Each event contract has a nominal value of $1, and the market price represents the probability of the event occurring; for example, $0.65 indicates a 65% probability. Kalshi calculates fees based on the expected return of the contract, with the core formula being:
Fees = 0.07 × C × P × (1 − P)
where C is the number of contracts traded, and P is the price. For some high-frequency or order book markets, the platform adopts a lower maker fee rate, and no fees are charged for unfilled orders. Kalshi does not take a share of user losses but operates on a traffic-driven revenue model centered on trading activity, fundamentally differentiating it from gambling platforms. In addition to trading fees, the platform charges service fees for debit card deposits (2%) and small withdrawals ($2), but ACH and bank wire transfers are free. Overall, this model ties Kalshi's revenue closely to market trading volume, providing predictability and sustainability.
3.1.2 Decentralized Representative Polymarket
Development History
Polymarket is a microcosm of the transition from concept validation to institutional infrastructure for decentralized prediction markets. Its story began in 2020, founded by New York entrepreneur Shayne Coplan, and launched its early version on the Polygon network. Unlike previous gambling-style prediction markets, Polymarket positioned itself from the outset as a public information aggregation platform, attempting to reflect collective cognition and event probabilities through price mechanisms. The platform employs a decentralized matching and clearing structure, without counterparties, using an order book mechanism to match trades between users, thereby minimizing information distortion and manipulation. In 2020, Polymarket received early investments from institutions like Polychain, completing product prototypes and initial user growth, gradually forming its prototype of event financialization in the crypto world.
However, decentralization did not shield it from regulatory impacts. In 2022, the CFTC determined that some of its event markets constituted unregistered derivatives trading, imposing a $1.4 million fine on the platform and requiring it to shut down access for U.S. users. In response to this setback, Polymarket chose a strategy of retreating to survive; it voluntarily exited the U.S. market while inviting a former CFTC chair to serve as an advisor to establish policy communication and compliance pathways.
After experiencing regulatory growing pains, Polymarket did not halt its expansion. In 2024, the U.S. presidential election cycle became a new growth engine for the platform, attracting significant funds and media attention to markets related to elections, inflation, macroeconomics, and technology events. Polymarket's trading volume and number of market creations reached historic highs, with daily liquidity increasing more than threefold compared to the previous year, and its probability curves were frequently cited by mainstream media. During this period, Polymarket evolved from a crypto-native application into a financialization outlet for real narratives, validating the communicative power and public opinion penetration of prediction markets.
At this stage, Polymarket has entered a true structural turning point. In July 2025, the U.S. Department of Justice (DOJ) and the Commodity Futures Trading Commission (CFTC) officially concluded their investigation into Polymarket. Subsequently, Polymarket acquired QCEX, a derivatives exchange and clearing company licensed by the CFTC, for approximately $112 million, and announced its return to the U.S. market in September 2025 (the specific launch date for the U.S. version has yet to be announced). In October of the same year, the parent company of the New York Stock Exchange announced plans to invest $2 billion in Polymarket, with a valuation reaching $9 billion. This series of initiatives marks Polymarket's transition from a crypto-native application to a compliant and institutionalized financial infrastructure, transforming its role from a prediction market platform to a probability data provider and information financial infrastructure, symbolizing the gradual integration of prediction markets into the core landscape of the mainstream financial system.
User Data
Currently, Polymarket's monthly trading volume is approximately $1.1 billion. At the same time, the monthly trading volume data of Polymarket clearly reflects its narrative and traffic cycles. During the 2024 U.S. presidential election, the platform's monthly trading volume continuously rose to a peak, becoming the most notable growth point among all crypto applications at that time. This explosive growth is not coincidental; on one hand, political events inherently carry strong uncertainty and attention; on the other hand, Polymarket provides a more intuitive probability price than traditional betting, allowing users to trade quickly in a real-time opinion environment.
Source: Dune, https://dune.com/rchen8/polymarket
In terms of user growth, it presents a typical exponential explosion curve. From its launch in 2020 to the end of 2023, the platform's user count remained low and stable, with monthly active users below 10,000. The real turning point occurred in early 2024, driven by hot events such as the U.S. presidential election and economic data predictions, leading to a sharp increase in new user numbers, with the total number of unique users exceeding 1 million for the year. In the chart below, it is evident that new users (green) and active old users (blue) formed a synchronous peak in 2024, indicating that the platform not only attracted a large amount of new traffic but also activated the activity of existing users. However, it is important to note that the platform currently has approximately 1.33 million cumulative active users, with an average daily user count of about 40,000, of which approximately 38,000 are daily active old users and about 2,000 are daily active new users. This indicates that the platform's user growth is slowing down, necessitating the development of new growth engines to expand the incremental user base and activate existing user activity.
Source: Dune, https://dune.com/rchen8/polymarket
From Polymarket's weekly nominal trading volume, similar to Kalshi, its market structure also exhibits a clear thematic concentration characteristic. Recent data shows that the three most active categories in trading are also sports, politics, and crypto. The sports category ranks first with approximately $273 million, demonstrating high participation during major sporting events; the political category follows closely with $256 million, maintaining strong interest and reflecting users' ongoing attention to global political events; the crypto market ranks third with $165 million, indicating that market speculation interest in crypto assets and their price trends remains robust. These three categories collectively account for about 80% of the overall trading volume, forming the main trading drivers for Polymarket.
In terms of category trading volume comparison, while both Polymarket and Kalshi exhibit the three dominant directions of sports, politics, and crypto, there are significant differences in structural distribution. Kalshi's market is highly concentrated, with sports trading volume reaching $818 million, about three times that of Polymarket, reflecting its users' preference for short-term, clearly defined event predictions; in contrast, Polymarket's structure is more balanced, significantly leading Kalshi in the political and crypto markets. Overall, Kalshi presents a centralized structure centered on sports, while Polymarket forms a diversified pattern of politics and crypto, reflecting the essential differences in participant demographics and risk preferences between centralized and decentralized prediction markets.
Source: Dune, https://dune.com/rchen8/polymarket
From Polymarket's global traffic sources, the U.S. remains its absolute core market, accounting for approximately 29.98%, a 42% increase from the previous period, primarily driven by the upcoming U.S. elections and the sports season. Canada (8.8%), Germany (6.1%), the UK (3.5%), and India (3.3%) follow, but all experienced declines in traffic ranging from 5% to 30%, indicating that user stickiness in markets outside North America remains relatively weak. In contrast, Kalshi's geographic distribution is more concentrated but deeper in growth. Its U.S. traffic share reaches 78.75%, indicating that it has established a solid institutional and retail foundation in the U.S. as a compliant event trading exchange. Although Kalshi's international penetration rate is still lower than Polymarket's, it has significant advantages in compliance channels, USD deposits, and institutional investor participation. Polymarket's traffic is broader but more dispersed, while Kalshi's traffic is more concentrated but deeper.
Source: Similarweb, https://secure.similarweb.com
Operational Mechanism
The core of Polymarket is event contracts, where users express their judgments on the outcome of an event by buying and selling "YES/NO" shares. The rules for each market are defined by the platform or community creators, including event content, verification sources, settlement time, and invalidation conditions. Buying a "YES" share bets that the event will occur, while buying a "NO" share bets that it will not. Share prices range from $0 to $1, representing the market's real-time estimate of the probability of the event occurring. However, this price is not set by the platform but is formed by real transactions in the order book. When the price a buyer is willing to pay matches the price a seller is willing to accept, the price at which the transaction occurs becomes the new market benchmark; the platform typically displays the midpoint of the bid-ask spread as the current price. For example, at $0.68, if the highest bid in the market is $0.66 and the lowest ask is $0.70, the system will display the midpoint of $0.68 as the current probability. If a buyer chooses to buy at market price, accepting the seller's $0.70 ask, the transaction price will be updated to $0.70, and new buy and sell orders will rearrange around this transaction price, with the price adjusting accordingly. Therefore, the $0.68 does not represent odds set by the platform but rather the equilibrium probability achieved under the constraints of real buying and selling forces in the market. Currently, Polymarket supports mainstream crypto wallets such as MetaMask, Coinbase Wallet, Phantom, and OKX Wallet. For users who do not use traditional crypto wallets, Polymarket allows trading through email or by registering an account to generate an internal proxy wallet. By depositing USDC through the Ethereum and Polygon networks, users can intuitively observe their betting situation on the dashboard page after placing orders.
Source: polymarket.com, https://polymarket.com/
Polymarket employs an on-chain order book matching mechanism and ensures market continuity through market makers and algorithmic pricing. Transactions are settled using stablecoins (currently mainly USDC), with price changes instantly reflecting market sentiment and capital flows. Unlike traditional betting, Polymarket does not act as a counterparty but only provides the infrastructure for matching, clearing, and information verification, so prices are entirely determined by participant behavior, forming a decentralized collective probability expectation. After each market reaches its specified date, the system verifies the results from trusted sources according to preset rules.
Business Model
Polymarket's business model differs from traditional exchanges. For a long time, the platform did not charge transaction fees for most markets, sacrificing short-term revenue to gain user base and liquidity depth. This zero-fee strategy is uncommon in the crypto industry, but Polymarket quickly seized the mental high ground in the prediction market. As the platform scales, Polymarket's potential revenue sources are becoming clearer. First, the commercialization prospects surrounding data and probabilities are widely optimistic. The platform's prices have been cited multiple times by mainstream media as reference indicators for event probabilities, and collaborations with data providers like X indicate that it is positioning market data as a new commercial outlet. Second, the U.S. version of the platform, set to launch after compliance, is expected to introduce a traditional exchange-style fee model, including matching fees, clearing fees, and API data subscriptions, thereby achieving sustainable cash flow. Meanwhile, Polymarket may also explore collaborations with brands and content providers, such as launching specialized markets around major sporting events or popular crypto narratives. Such collaborations could not only bring direct sponsorship revenue but also further enhance its external communicative power.
Ecosystem Landscape
As of now, Polymarket's ecosystem has expanded from a single prediction platform to a complete information financial network covering trading, AI agents, data analysis, and social dissemination, encompassing a total of 14 categories and 118 projects. On the trading front, peripheral protocols and tools built around Polymarket, such as betmoardotfun, are forming diversified event trading entry points, covering automated trading, arbitrage aggregation, and strategy replication functions, providing high-frequency liquidity and speculative depth to the market; in the smart agent domain, over a dozen projects focused on AI prediction, model pricing, and probability generation, such as trypolyagent, have emerged in the ecosystem, integrating with Polymarket via APIs or plugins to offer users automatic ordering and probability modeling services, driving the platform towards AI-driven predictive infrastructure; in terms of data and analysis, dashboards and tracking tools from institutions like blockworksres have achieved real-time tracking of Polymarket's trading volume, thematic popularity, and capital structure, forming the market analysis layer; meanwhile, social and alert applications like PolyAlertHub have formed a peripheral traffic layer, responsible for information dissemination and the diffusion of public opinion momentum.
Source: @PolymarketEco, https://x.com/PolymarketEco
At the same time, Polymarket is continuously expanding both horizontally and vertically in infrastructure. Currently, Polymarket supports the HyperEVM network, and it is advancing deep integration with MetaMask, allowing users to conveniently connect their wallets, deposit funds, and participate in prediction markets. In the future, it may be possible to create and trade prediction markets directly within the wallet interface without needing to switch platforms.
3.1.3 Other Integrations
3.1.3.1. Limitless
Limitless is a decentralized prediction market based on the Base blockchain, adopting a model closer to traditional exchanges by implementing a complete central limit order book system on-chain. Limitless allows users to place orders, match trades, and cancel orders just like in centralized exchanges, supporting market orders and limit orders, with prices determined by supply and demand from both sides of the market. For multi-option events, the platform introduces a multi-directional contract design, enabling users to configure positions for multiple outcomes simultaneously and flexibly close positions before the event concludes, locking in profits or losses in advance.
This interaction logic is very intuitive for users with high-frequency trading experience while retaining the transparency and verifiability of on-chain operations. Limitless leverages the high performance and low latency of Base to enable high-frequency matching, while utilizing the decentralized oracle Pyth for instant price settlement, reducing the risk of market manipulation. Limitless focuses on short-cycle markets, offering hourly and daily prediction contracts covering themes such as Bitcoin and Ethereum price fluctuations, U.S. stock indices, and macroeconomic indicators. Users connect to the Base chain through Web3 wallets and can also provide liquidity by placing limit orders or invite newcomers to earn points.
In addition to the points incentive mechanism, Limitless has also introduced an LP reward mechanism, incentivizing limit order placers with USDC. As shown in the figure below, this mechanism requires that the order size exceeds a minimum threshold to be included in the midpoint calculation and reward range. The reward amount is determined by three parameters: the total daily amount, the maximum eligible price spread, and the minimum order quantity. The closer the order is to the market midpoint and the larger the size, the higher the reward received, with the system calculating rewards on a minute-by-minute basis and settling daily. This mechanism effectively prevents small orders from manipulating prices or gaming rewards, guiding LPs to provide meaningful market depth and tight spreads, thereby enhancing liquidity and market efficiency.
Source: limitlesslabs, https://limitlesslabs.notion.site
Limitless's exploration indicates that the combination of high-performance on-chain trading and prediction markets is expected to attract professional traders, introducing new liquidity and gameplay to the sector. Currently, Limitless has completed two rounds of financing, totaling $7 million. In September 2024, it raised $3 million in a seed round, with investors including 1confirmation. In July 2025, Limitless completed a strategic financing round of $4 million, with investors including Coinbase Ventures and 1confirmation. According to Dune data, Limitless's cumulative trading volume has exceeded $320 million, and its total monthly user count has surpassed 200,000, making it the fastest-growing prediction market on the Base chain.
Source: Dune, https://dune.com/limitlessexchange/limitless_
3.1.3.2. MYRIAD
As a representative of embedded prediction markets, Myriad is dedicated to integrating predictions into content consumption scenarios. Myriad is based on the Abstract chain and incubated by the DASTAN team (the parent company of Decrypt media), which has deep media resources. It enables users to bet on what they read through a browser plugin, presenting relevant prediction market opportunities directly while browsing news, X, and other content, allowing for one-click betting.
In terms of promotion, Myriad has already conducted integration pilot projects with emerging media platforms like Rug Radio. For example, Myriad's prediction interface is embedded in Rug Radio's community content, encouraging readers to bet and discuss related news events. Myriad's innovation lies in transforming predictions into social interactions, allowing more non-professional users to participate in prediction markets. Although this requires users to change their reading habits, once an atmosphere is formed, it will greatly expand the user base of prediction markets. Myriad can also accumulate a large amount of user behavior and opinion data, feeding back into content production and distribution, achieving multiple benefits. Currently, Myriad is gradually opening more thematic markets and plans to introduce more media partners. If its model succeeds, we may see prediction markets become standard features on media websites and forum communities, subtly completing user education and market expansion. As of now, Myriad's total trading volume is $14.66 million, with monthly transaction counts reaching 4.16 million. Compared to the trading volumes of other prediction markets, Myriad still has significant growth potential.
Source: Dune, https://dune.com/surfquery/myriad-markets
3.2. DeFi Category
3.2.1. Gondor
Gondor is building a DeFi infrastructure layer for prediction markets, aiming to address the low capital utilization issue in prediction markets. Currently, funds participating in prediction markets are locked for long periods. Gondor's solution is to launch a position collateral lending protocol, allowing users to use their existing prediction market positions (initially supporting Polymarket) as collateral to borrow stablecoins like USDC from the platform. The positions users collateralize are essentially assets that fluctuate with market prices, and their value adjusts in real-time with changes in prediction market prices, settling based on results at the end of the event. Therefore, users still bear the price risk of their original prediction positions but can borrow some funds from Gondor for other DeFi scenarios without interrupting their prediction exposure. For example, if a user bets $1,000 USDC on Trump winning on Polymarket, they would traditionally have to wait for the result to be revealed before using the funds again; however, through Gondor, that position can be collateralized to borrow about $500 USDC for other DeFi scenarios without affecting the original bet, significantly improving efficiency and reducing opportunity costs.
Source: @0xkeshav, https://x.com/0xkeshav
In August, Gondor announced the completion of an angel round of financing with participation from institutions like Maven11 Capital and reached a strategic cooperation intention with Polymarket. Currently, the Gondor Beta version is in a small-scale testing phase, and users must complete KYC through a 20-minute video session to participate. Gondor's long-term vision is to become a lending pool for all prediction markets, not only serving Polymarket but also planning to support platforms like Kalshi in the future. It is foreseeable that if Gondor's model matures, it will attract more rational investors to join prediction markets, as they will no longer worry about locked liquidity and can simultaneously earn interest or mining rewards. This will bring hedging funds into prediction markets, further driving their trading volume growth. Therefore, Gondor is expected to become a new development direction for prediction market infrastructure.
3.2.2. Melee
Melee is an innovative project based on Solana, described as a combination of Pump and prediction markets, with the core idea of activating market liquidity through meme-ified prediction events. Unlike traditional prediction markets that rely on platform teams to set questions and introduce market makers to provide liquidity, Melee adopts a completely decentralized and permissionless structure, allowing anyone to create prediction markets and issue corresponding meme tokens on any topic. This design breaks the boundaries of prediction markets, combining the openness of the internet with crypto speculation culture, giving rise to a new market form, also known as "viral markets." On Melee, prediction events are no longer just financial tools for betting correctness but are packaged as meme tokens with social attributes and speculative potential. For instance, if a user creates the event "Will a certain internet celebrity have over 10 million followers by the end of next year?", they can issue the token $FAN and set different price ranges as prediction outcomes. Investors optimistic about the event will buy $FAN to drive its price up, while those who are not will choose to sell or short. Upon the event's expiration, the system determines the final result based on the price range in which $FAN falls.
Melee's similar social mechanism allows prediction events to gain sustained attention after becoming memes, thereby bringing new incremental funds to the entire prediction market. Especially for long-tail events that mainstream platforms often do not list, social dissemination is more suitable for generating trading momentum; KOLs, community discussions, and topic diffusion will naturally attract speculators and retail participants. As a result, long-tail events will no longer remain dormant due to a lack of platform support or liquidity but will rely on social networks to form a self-circulating drive of funds and emotions, enriching the trading scenarios and ensuring continuous capital inflow for the prediction market as a whole. Melee's design simultaneously meets the needs of three core groups:
Market Creators: They can issue meme tokens around hot topics, cultural events, elections, sports, or entertainment issues, directly monetizing attention. Unlike traditional prediction markets, creators do not need to bear the reputational risk of outcome deviations; they profit solely through community size and trading volume.
Traders and Speculators: Thanks to the meme-ification mechanism, early participants can buy tokens at extremely low costs, and once the market heats up, they can reap substantial returns through price increases.
General Users: They can gain interactive and entertainment value during the real-time fermentation of cultural events, combining discussions about events with speculative behavior, creating new participation motivation.
As of now, Melee is still in a closed testing phase, with a waiting list exceeding 25,000 people. The team has completed $3.5 million in financing, with investors including Variant and several strategic angel investors, with backgrounds from leading tech and financial institutions such as Solana, Avalanche, Monad, SIG, Microsoft, and Amazon. This provides solid support for the project's future expansion. Notably, Variant has also invested in well-known projects like Uniswap and Morpho.
Source: @jessewldn, https://x.com/jessewldn
In a longer-term vision, Melee aims to become a platform for pricing human beliefs. Through the mechanism of "viral markets," millions of markets will be created and resolved in real-time globally, with market prices reflecting the immediate value of cultural trends, social beliefs, and entertainment events. With the rise of trend trading platforms like Noise, Melee is expected to become an important connection point between prediction markets and meme trading ecosystems, where the former is responsible for packaging predictions into tradable memes, and the latter drives the broader dissemination and trading of these memes. The combination of both could give rise to a new financial ecosystem centered around attention. Melee's breakthrough lies not only in technological or product model innovation but also in its redefinition of how prediction markets can be popularized, entertained, and integrated with crypto speculation culture. If this model succeeds, it could redefine the boundaries of prediction markets and enable them to achieve internet-scale expansion.
3.2.3. Flipr
Flipr initially positioned itself as a social entry point for prediction markets, with the core goal of making prediction trading as simple as posting. It is based on Polymarket's event contracts, allowing users to place bets, open positions, or close positions directly on X by tagging @fliprbot. Users can trade without paying gas fees, making the operation almost barrier-free. It can also be understood that users can simultaneously establish prediction positions while discussing events. This design allows prediction markets to break away from traditional DeFi interface logic and shift towards social dissemination logic, accumulating an initial user base and liquidity for Flipr.
At the same time, as prediction markets further develop, Flipr has found that the infrastructure for prediction markets remains at the spot stage, lacking leverage, risk control, and a unified trading terminal. To address this, Flipr has proposed a new positioning, launching an aggregation terminal and a self-leveraged model, aiming to become a provider of infrastructure for prediction markets. The system is designed to avoid the "instant liquidation" problem of perpetual contracts by enhancing risk resilience through layered collateral (base margin and refundable insurance), volatility protection, and a weighted average price settlement mechanism. Users can directly call Flipr's terminal interface to execute trades on Polymarket and Kalshi's event markets, with the system supporting features like take-profit and stop-loss, liquidation warnings, and one-click arbitrage.
Its goal is to become a combination of Hyperliquid and Jupiter for prediction markets, providing efficient leverage mechanisms at the upper layer while integrating cross-platform liquidity at the lower layer. If its leverage model can successfully operate in low liquidity environments, Flipr has the opportunity to become an important infrastructure for the specialization and institutionalization of prediction markets. However, it is worth noting that the Flipr team is still at an initial scale, currently consisting of only two full-time and two part-time members, and the entire team operates anonymously. Previously, Flipr was also caught up in a "FUD" controversy, but the official response denied the "FUD" incident. Meanwhile, the project's operational pace and product iteration have been relatively slow, indicating that Flipr requires long-term observation.
3.3. AI** Category**
3.3.1. Opinion
Opinion is based on the BNB Chain and Monad, receiving early investment from YZi Labs in 2024 and completing a $5 million Pre-Seed round in March 2025, with investors including YZi Labs, Amber Group, and Animoca Ventures. The core goal of Opinion is to reconstruct the underlying logic of prediction markets, making "opinions" a priceable digital asset and achieving deep integration between the social layer and the prediction layer.
By binding predictions with expression, trading with social interaction, Opinion has created an "Opinion-to-Earn" system, where opinion producers are both participants in the market and creators of value. In terms of system design, Opinion Labs' product mechanism focuses on three directions: stability, efficiency, and openness, with the core aim of addressing the structural bottlenecks present in traditional prediction markets.
In terms of underlying architecture, Opinion has reshaped the trading and funding logic of the market by combining on-chain limit order books (CLOB) with Meta Pool liquidity pools. The CLOB module ensures transparency and traceability in the trading process, while the Meta Pool aggregates idle funds across markets, enabling fund reuse and deep enhancement. The synergy between the two allows price discovery to return to each real trading action, significantly reducing slippage and greatly enhancing depth.
In terms of settlement mechanisms, Opinion deeply embeds AI into the market operation logic, constructing a hybrid model of AI prediction and community review. AI is responsible for front-end event identification and invalid market filtering, ensuring the timeliness and quality of prediction topics; at the same time, during the settlement phase, AI provides preliminary judgments, which are then reviewed and confirmed by an on-chain jury, forming an efficient and fair two-tier arbitration process.
In terms of market positioning, Opinion focuses on the macroeconomic prediction field, placing emphasis on key economic indicators such as CPI, employment data, and interest rate decisions, distinguishing itself from other platforms that primarily focus on short-cycle markets related to political and entertainment events. This strategy makes prediction markets more aligned with real economic activities, allowing users to express their judgments on macro trends through trading, letting market prices naturally settle into collective consensus.
Source: Opinion Labs, https://opinionlabs.medium.com/
The Opinion Labs team is strong, with core members from top financial and tech institutions such as JPMorgan, McKinsey, and Amazon, possessing a composite background in quantitative trading, data science, and AI model construction and development. Currently, Opinion has attracted approximately 1.61 million users to participate in the testnet phase, with a cumulative trading volume of 190 million USDO (test tokens), covering 633 markets across diverse fields including politics, crypto, sports, and community. Its mainnet whitelist is now open, and users can fill in their email addresses for a chance to participate early. Opinion is expected to establish a first-mover advantage in the AI-driven prediction market space and become an important infrastructure in the direction of financializing opinions.
3.3.2. Talus
Talus is an infrastructure project focused on the combination of AI agents and prediction markets. Talus believes that AI agents not only execute tasks and produce content but also generate predictable and bettable events in competition. These agents can operate independently or compete against each other in prediction markets, creating new forms of investment and entertainment. Since 2024, Talus has completed multiple rounds of financing. In February 2024, Talus secured $3 million in a seed round led by Polychain Capital, and in November of the same year, it completed another round of $6 million financing, with a valuation of approximately $150 million. In September 2025, Talus received strategic investments from the Sui Foundation and Walrus Foundation, focusing on supporting Talus's foundational deployment in the Sui chain ecosystem and decentralized storage integration.
In terms of prediction markets, Talus introduces an Agent vs Agent (AvA) competitive mechanism, transforming the adversarial process between AI into bettable market events. Meanwhile, ordinary users, like spectators in sports betting, can choose which AI to support before the competition starts and place bets on the predicted outcomes on-chain. After the competition ends, the system automatically settles the earnings based on the real results recorded on-chain. This model allows the behavior and decisions of AI to become a new asset class, with each confrontation and each match being an independent prediction market event. Its first application product, Idol.fun, has launched on the testnet, allowing users to publish social content, generate responses from different AI agents, and place bets on the performance of the agents based on interaction heat or public feedback, thus realizing a gameplay of predicting AI decisions. According to official plans, Idol.fun is set to officially launch on the mainnet in the first quarter of 2026 and will become one of the first AvA game entrances centered around predicting artificial intelligence. In this system, AI agents are both participants and creators of assets; the changes in market prices become the on-chain expression of the strength of AI competition.
Source: @idoldotfun, https://x.com/idoldotfun
The core challenge facing Talus is how to ensure that AI behavior is truly priceable and trusted in the market. While the competition between AI agents can be efficiently completed on-chain, if there is a lack of transparent verification of results and stable anchoring of economic value, the entire prediction could devolve into short-term gaming. Talus attempts to establish a trust foundation for "behavior as an asset" by fully putting the decision-making process, interaction records, and result data of agents on-chain through the Nexus protocol and Walrus auditing system. Meanwhile, the tokenization mechanism planned for the project will combine prediction incentives, agent earnings, and market liquidity, making AI behavior not just a product of model computation but forming a continuous value closed loop that is assessed, priced, and traded in the market. If this mechanism can operate stably, Talus is expected to bring structural reshaping to prediction markets, moving from a reliance on external event gaming to an endogenous prediction system driven by the behavior of agents, thus enabling prediction markets to truly possess the ability to self-generate events and self-expand.
3.4. Other Categories
3.4.1. Polyfactual
Polyfactual is one of the earliest projects in the prediction market ecosystem to receive the official badge from Polymarket, attracting the attention of the CEOs of both Polymarket and Kalshi. Currently, Polyfactual has achieved over $230,000 in revenue by live streaming political and global events weekly and interpreting them through prediction markets, proving the commercial value of information entry in the prediction space. However, Polyfactual's goal is not just to become a content media outlet but to build a true market-making and risk management infrastructure within the prediction market.
Its core plan is divided into two main lines. The first is Project X, with its Beta version scheduled to launch in November 2025. The product's token system is built on the TBT mechanism. Unlike traditional speculative tokens, TBT is a "reinsurance-type asset token" supported by a real capital pool, with each TBT anchored to the reserve funds held by the platform. Users purchasing TBT essentially provide risk protection for the entire prediction market: a portion of the funds flows into an insurance pool to execute payouts in case of settlement delays, dispute determinations, or extreme volatility on platforms like Polymarket. This mechanism effectively introduces a risk mitigation layer similar to central clearinghouses in traditional finance, allowing market makers to continue providing deep liquidity during periods of volatility. The second is Project Y, which focuses on automating cross-platform arbitrage, aiming to capture price discrepancies and efficiency gaps between Polymarket and Kalshi. Through real-time monitoring and automated execution arbitrage bots, the system locks in risk-free price differences in illiquid markets and proportionally distributes the profits to $POLYFACTS token holders. This design not only enhances the overall capital efficiency of the prediction market but also injects sustainable cash flow into the Polyfactual ecosystem.
Source: Polyfactual, https://www.polyfactual.com/
The $POLYFACTS token is issued on the Solana chain, with its token functions tied to product revenue, governance, and community engagement. Holders can not only directly share in the protocol's profits but also gain priority access to the Project X Beta phase and its official launch upon meeting certain conditions; in the future, they will also participate in topic selection and directional voting for live content. The logic of Polyfactual is to combine market making, insurance, and arbitrage mechanisms with narrative entry points, creating a triple logic for prediction markets. Live content attracts external traffic into the market, Project X reduces the risk exposure of liquidity providers, and Project Y enhances cross-market pricing efficiency while rewarding token holders. If these modules are successfully implemented, Polyfactual will play a dual role as both a public opinion entry point and a liquidity hub in the prediction market ecosystem, bringing the trading logic of prediction markets closer to a sustainable market-making model found in traditional capital markets.
3.4.2. Noise
Noise is a platform that transforms narrative heat into tradable assets, with the core logic being to allow users to trade attention trends directly, rather than specific tokens or event outcomes. Noise is currently primarily built on the Ethereum Layer 2 MegaETH, creating a series of topic indices based on on-chain contracts, allowing users to long or short specific narratives with up to 5x leverage, thus betting on the direction of market sentiment changes. Essentially, Noise has built a market for attention derivatives, abstracting common narrative cycles, FOMO psychology, and public opinion fluctuations in the crypto market into tradable price signals, enabling traders to profit from the warming or cooling of stories. The Noise testnet is now live, initially opening multiple long and short trading options for various narrative indices. However, its team is still relatively small and in the prototype development stage, with platform functions centered around heat index trading and not yet offering complex derivatives or governance mechanisms.
The core of Noise lies in reshaping the boundaries of prediction markets, from event outcomes to emotional trends, from asset prices to attention flows. As its tradable indices expand to broader information dimensions, including market shares of platforms like Kalshi, on-chain ecological narrative heat, brand or topic influence, etc., Noise is attempting to systematically financialize market attention. This means that in the future, investors can build portfolios based on topic heat, hedge narrative risks, and even position themselves ahead of "stories yet to land." If the platform can achieve breakthroughs in the authenticity, anti-manipulation, and liquidity depth of heat data, it may become a new narrative for prediction markets.
Source: @noisexyz, https://x.com/noisexyz
Football.Fun is an on-chain sports prediction platform running on Base, replacing traditional luck-based betting with skill-based gameplay. Users can purchase on-chain shares of players, with each player corresponding to a tradable digital asset whose price changes in real-time based on the player's performance in matches, team results, and market expectations. Unlike traditional football or betting platforms, Football.Fun is designed to allow players to truly own the players they choose; these assets will not be cleared or invalidated due to the end of the season, game updates, or copyright issues, allowing users to retain, trade freely, or wait for market repricing in the future. This mechanism allows fans to be not just participants but also investors holding their favored sports assets. Currently, the platform has a total TVL of $10.54 million, total revenue of $3.61 million, and total trading volume exceeding $60 million, with a cumulative active wallet count of 14,026.
Source: Dune, https://dune.com/tervelix/fdf-footballdotfun-mega-dashboard-terminal
The main innovation of Football.Fun lies in splitting each player into tradable shares on the platform, with prices adjusted in real-time by smart contracts. Players can build their portfolios based on recent player status, team tactics, and other information, forming trading decisions based on cognition and analysis. This structure makes sports predictions closer to an on-chain intellectual competition rather than a simple odds game. The platform also features a points and card pack reward system, allowing players to accumulate skill points through wins and trades, which will affect player rankings and reward distribution.
The rise of Football.Fun reflects a clear trend: sports prediction is shifting from an odds-based economy to a fan sovereignty economy. It centers not on betting odds but on fan judgment as a value anchor, transforming emotional engagement into asset liquidity. The platform plans to launch a native token, $FUN, which will be used for trading fee sharing and community governance in the future. For users, this means that predictions are no longer about betting on outcomes but about holding positions based on opinions. The emergence of Football.Fun marks the beginning of on-chain sports financialization, with a new asset layer driven by information cognition and fan consensus rapidly taking shape outside the sports arena.
## IV. Data and Market Structure
4.1. Market Overview
According to data from Defillama, the total locked value of prediction market protocols is approximately $225 million. As observed in the chart below, the total locked value of prediction market protocols remained low between 2020 and 2023, fluctuating between approximately $10 million and $50 million. However, since mid-2024, with the surge in trading volume on platforms like Polymarket due to the U.S. elections, the total locked value has experienced a structural breakthrough, climbing to a peak of $500 million at the beginning of this year, before retreating and stabilizing above $200 million.
Source: Defillama, https://defillama.com/yields/stablecoins
On the other hand, as mentioned earlier in this report, the ecological landscape of prediction markets is showing trends of thematic diversification and narrative stratification. Prediction markets are gradually evolving from being "single-event driven" to "multi-narrative driven," with the market structure and user demand maturing and becoming more long-term.
4.2. User Analysis
The user structure of prediction markets exhibits a clear polarization. On one end are information-driven users, who are generally older and more educated, participating in predictions primarily to obtain information and express opinions. These users are often active in political and macro markets on platforms like Polymarket, viewing predictions as tools to validate personal judgments and gain insights into public sentiment. Their trading frequency is low, but they maintain long-term attention on specific topics. On the other end are speculative arbitrage users, who view prediction markets as a new channel for short-term profits, exhibiting behavior more akin to high-frequency traders. For example, many users on trend trading platforms like Noise come from the crypto circle, familiar with on-chain leverage and hedging operations, often profiting from market volatility and information lags. These users are more focused on profit efficiency and capital utilization, making them the main contributors to platform liquidity and trading depth.
Due to regulatory restrictions and regional differences, there are significant disparities in the platforms accessible to users in different regions. U.S. users primarily access Kalshi and Polymarket, while the acceptance of prediction markets is rapidly increasing in the crypto communities of Europe and Asia. In the long term, regulatory trends will directly impact user composition and platform growth paths. If the U.S. CFTC relaxes restrictions on event contracts in the future, compliant platforms like Kalshi may experience explosive growth aimed at the general public; conversely, if decentralized platforms face stricter crackdowns, some speculative users may shift to underground markets or legal platforms, resulting in a new user re-segmentation.
In terms of user retention rates, prediction markets are generally influenced by event cycles, exhibiting characteristics of surges during events followed by declines afterward. For instance, during the 2020 U.S. elections, a large influx of new users entered Polymarket, but activity quickly dropped post-election; a similar trend is expected in 2024, although the presence of airdrops and token issuance expectations has improved retention this time. Improving retention rates has become a core challenge for various platforms. Different projects are extending user lifecycles through incentive mechanisms, content expansion, and social design. For example, Limitless has introduced a points season incentive, and Flipr employs a social mining mechanism that allows users to earn points through trading and interaction, creating a sense of ongoing purpose; some platforms maintain market freshness and discussion heat by expanding themes such as sports, entertainment, and technology. Overall, the user structure of prediction markets is shifting from event-driven participants to long-term thematic traders, with platform competition focusing more on sustained user relationships and behavioral retention rather than single-event traffic.
4.3. Major Categories
There are many themes in prediction markets, but based on trading volume, user attention, and platform strategies, the most influential categories remain political, sports, finance and crypto, social entertainment, and macro social events. The following are the specific categories:
Political Category: This includes election results, whether policies are passed, trends in international relations, etc. Among these, election predictions attract the most attention and have the highest trading volume. Major elections often lead to peaks in the prediction market. The characteristics of the political market are strong cyclicality, binary outcomes, and highly public information. For professional investors, the political market can also be used to hedge policy risks; for example, hedge funds can bet on a particular party's victory to hedge against the impact of policy changes on the stock market. However, legal regulations impose the most restrictions on political markets. If policies loosen in the future, political prediction markets have the potential to grow several times and attract institutional funds.
From the weekly trading volume data of political prediction markets, Polymarket has long dominated the political event sector, with a market share of about 75%–85%, while Kalshi maintains around 15%–20%. Since the start of the U.S. election cycle in the fourth quarter of 2024, political predictions have become the core growth engine of the entire sector. Polymarket has repeatedly set weekly trading records at key moments, such as Biden's withdrawal and Trump's nomination, thanks to its relatively open event creation mechanism and social dissemination effects. In contrast, while Kalshi has compliance advantages, its event coverage is more conservative, focusing mainly on official elections and macro policies, resulting in a relatively stable trading structure. Overall, the political category exhibits a clear winner-takes-all pattern, with Polymarket driving growth through social traffic and an open ecosystem, while Kalshi relies on compliance and institutional users to maintain stability.
Source: Dune, https://dune.com/datadashboards/prediction-markets
Sports Category: This includes outcomes of events such as football, basketball, and rugby, league championships, player statistics, etc. Sports prediction markets have been particularly active in recent years, becoming a new growth force. However, sports contracts still face challenges on-chain, including compliance, reliability of data sources, odds design, and gameplay innovation. Traditional bookmakers and centralized sports betting platforms have inherent advantages in user habits, brand penetration, and regulatory permissions. For on-chain platforms to win the competition, breakthroughs in unique gameplay or low latency and high experience are needed. If a sports prediction platform becomes a "killer app" on-chain in the future, the proportion of sports could continue to rise.
From the weekly trading distribution of sports prediction markets, Kalshi has a significantly larger share, maintaining about 70%, while Polymarket is around 30%. Kalshi quickly accumulated depth in this field within a few months of launching sports contracts, establishing its dominant position in sports contracts. In contrast, Polymarket relies more on community and on-chain momentum for its sports contracts; while its trading is volatile, it has consistently remained in a secondary position in overall share. Due to the high frequency, broad audience, and easy dissemination characteristics of sports contracts, they have become a key battleground for both platforms; Kalshi's lead in this area reflects its structural advantages in market-making capability, capital capacity, and institutional liquidity.
Source: Dune, https://dune.com/datadashboards/prediction-markets
Finance and Crypto Category: This category includes predictions on cryptocurrency prices, macroeconomic indicators, and financial market trends. It overlaps to some extent with traditional derivatives, but prediction markets can offer more flexible and non-standardized underlying assets. Additionally, prediction markets still have application potential in hedging risks. However, this category is also facing competitive pressure. With the maturation and popularization of decentralized options, perpetual contracts, futures, and other financial derivatives, users may prefer to directly engage with the market through these tools. To maintain attractiveness, prediction markets need to differentiate themselves in designability, low barriers to entry, and combinatorial logic.
From the weekly trading volume data of crypto prediction markets, Polymarket holds an overwhelming advantage in this field, with its trading volume accounting for about 75%, while Kalshi's share is around 25%. This gap reflects the essential differences in ecological positioning between the two; Polymarket is closer to crypto-native users, covering high-volatility topics such as Bitcoin prices and spot ETF approvals, making it easy to attract on-chain speculators and information traders. At the same time, Polymarket's on-chain structure allows for higher settlement speed and transparency, enhancing the participation stickiness of retail and community users. In contrast, while Kalshi offers some macro and crypto-related contracts, its product design is more conservative and less volatile due to compliance restrictions, with trading depth mainly coming from risk hedging and less speculative accounts.
Source: Dune, https://dune.com/datadashboards/prediction-markets
Social Entertainment Category: This includes film awards, reality show competition results, celebrity gossip, weather changes, Nobel Prize allocations, etc. These topics are quite popular in traditional prediction markets but have a smaller proportion in crypto prediction markets. On one hand, crypto users are less concerned about these topics compared to political and financial ones; on the other hand, on-chain oracles find it less convenient to obtain results for these topics than for sports and financial data, and there is a larger space for controversy. However, emerging platforms like Myriad, which mention supporting predictions combined with media, indicate that this sector still has opportunities.
Macro Social Events Category: This category includes topics such as disease spread, climate change extent, whether technological breakthroughs occur, and the probability of major public events. It has certain public significance and research value and has been used for predictions in academic and public policy contexts. However, in crypto prediction markets, such contracts remain almost marginalized. Reasons include high verification difficulty, long result timelines, significant controversy, and strong uncertainty in return cycles. If more reliable data sources, AI model integration, verification program designs, or institutional collaboration mechanisms are established in the future, these contracts may hold strategic value between research, government, and markets.
Overall, currently, the sports and crypto categories are rising as the core engines of trading volume and activity; the political category remains a flagship category, but its growth rate is limited by regulations and cycles; the entertainment and macro social events categories remain as potential supplementary sectors. If the regulatory environment improves or breakthroughs in technology and gameplay innovation occur, the proportions among these categories may be reshuffled.
## Development Path and Potential Expansion
For prediction markets to achieve sustained growth, the key lies in forming synergistic effects with other on-chain fields. Currently, DeFi, AI, social integration, and protocol aggregation innovations are the most promising fusion directions. These directions can bring funds and users to prediction markets while also feeding back into their mechanism improvement and data accumulation.
5.1. Integration with DeFi
The integration of prediction markets with DeFi represents a shift from being a betting platform to becoming an on-chain financial infrastructure. In the past, "outcome shares" in prediction markets could only represent the direction and amount of user bets, but in the future, these shares may be standardized into freely tradable token assets that can be used across different protocols. The foundation for this change is the maturity of on-chain settlement mechanisms. In September 2025, the prediction platform Polymarket announced its integration with Chainlink, bringing decentralized data sources into the market settlement process, allowing objective events such as prices and sports results to automatically trigger settlements without manual voting. This significantly improves settlement efficiency and credibility, making prediction certificates easier to be accepted by DeFi protocols as composable collateral assets.
At the same time, the financialization of prediction markets is gradually being realized through new DeFi protocols, among which Gondor, mentioned earlier in this report, is particularly representative. Gondor is building a lending infrastructure for prediction markets, aiming to solve the structural issues of "long-term capital lock-up and low utilization." This means that participants in prediction markets can not only continue to hold risk positions but also invest the funds obtained from collateral into other DeFi scenarios. With the entry of institutional and hedge funds, such collateral lending protocols are expected to become important hubs for prediction markets to move towards DeFi.
5.2. AI Driving
The rise of AI may reshape the operational logic of prediction markets. Traditional markets rely on human participants to provide information and judgments, while AI can help the market price future events more quickly through large-scale data training and real-time information capture. The combination of prediction markets and AI mainly manifests in three aspects: information generation, automated trading, and human-machine collaboration.
In terms of information generation, AI models can continuously scan news, social media, and data sources to extract events with predictive value, then generate corresponding markets. Some projects have already attempted to use AI to automatically create "hot prediction topics," allowing the market's update speed to far exceed manual editing; in terms of automation, AI agents can act as market participants to execute orders, rebalance, and take profit. They can automatically adjust positions based on data changes, enhancing market depth and price response speed; most notably, AI and prediction markets form a "two-way flywheel." AI models obtain real-world probability feedback through market prices, continuously correcting their own understanding; while prediction markets enhance efficiency and trading volume through AI's high-speed reading and writing and algorithmic liquidity.
5.3. Social Extension
The integration of social networks transforms prediction markets from isolated financial tools into social behaviors that can be embedded in daily discussions. Projects like Flipr are attempting this model, allowing users to post bets or trade prediction positions directly on X using natural language, with each prediction operation linked to social interactions. Other users can choose to follow or bet against, creating a real-time feedback loop of opinions, emotions, and trades. This fusion of social and trading makes prediction markets not just financial operations but a new arena for information dissemination and collective emotional pricing.
The combination of prediction markets and social media is essentially the visualization of opinion trading. When users see a piece of news or an opinion on social platforms, the system simultaneously displays the current prediction odds, making probabilities an instant thermometer of public opinion. This design stimulates network effects; the more discussions there are, the more active the market becomes; the more prices fluctuate, the wider the social dissemination. However, social integration also brings new risks. When public opinion or misinformation spreads rapidly, prediction prices may be manipulated or misled in a short time. To balance the speed of dissemination and the credibility of results, some platforms have begun to introduce reputation mechanisms, delayed settlements, and AI-assisted judgment systems. In the long run, social media will become a traffic entry point and emotional amplifier for prediction markets, but stable mechanism design remains key to transitioning from virtual to real.
5.4. Protocol Aggregation
As the scale and event density of prediction markets rapidly increase, there has been a significant fragmentation of information flow. A single event is often simultaneously created on multiple platforms such as Polymarket and Kalshi, with different market odds, trading depths, and settlement standards, making it difficult for users to obtain optimal pricing and liquidity in real-time. The high-frequency triggering of news events further amplifies this contradiction, requiring predictors to switch between multiple platforms to complete information comparisons and order decisions. Just as the wave of meme coins has driven the emergence of trading aggregators and quantitative tools, prediction markets are now entering a new cycle where information overload leads to the rise of aggregation tools. Aggregators provide traders with cross-platform pricing and signal aggregation capabilities by integrating order books, identifying similar markets, automatically comparing prices, and analyzing news event correlations, reshaping the efficiency of the path from information to decision-making.
Currently, aggregation innovation projects are still in the early stages. However, in the long term, aggregators are expected to become the traffic hubs of the prediction market ecosystem, taking on roles similar to DEX aggregators in DeFi or meme market analysis platforms. With the entry of cross-platform arbitrage, AI strategies, and institutional funds, market participants are becoming increasingly specialized, and the core competitiveness of prediction markets may shift from the traffic advantages of a single platform to the comprehensive strength of the entire ecosystem in data response speed and capital utilization efficiency.
5.5. Bridging TradeFi
Currently, Polymarket has launched a U.S. stock prediction market in October this year. Users can directly trade on-chain events such as "Will Tesla's year-end stock price exceed $300?" transitioning from political and crypto predictions to traditional asset price predictions. This step shifts prediction markets from niche information games to real financial pricing logic, indicating that dollar-level liquidity is entering on-chain. The previously mentioned investment of $2 billion by ICE (the parent company of the New York Stock Exchange) in Polymarket may also be a bet on this trend, as prediction markets are no longer just a thermometer of public opinion but a potential foundational layer for on-chain derivatives.
Structurally, the U.S. stock prediction market is essentially on-chain binary options. This mechanism is extremely simple, transparent, and highly certain, making it almost a natural form of binary options on-chain. In the past, such products often appeared in gray OTC or centralized platforms, with opaque settlements and dispersed liquidity; Polymarket, however, automatically settles through smart contracts, making odds and liquidity pools completely public, giving on-chain binary options real financial attributes and transparent pricing logic. Prediction markets are blurring the lines between gambling and derivatives, as users betting on events are equivalent to buying "on-chain short-term options." As more traditional assets are introduced, prediction markets may evolve into on-chain micro-derivatives exchanges, becoming an important channel for the migration of TradFi liquidity on-chain.
5.6. Reshaping Attention
In the meme-dominated era of dissemination, attention was once the core traffic asset, but its spread relied on emotional resonance, lacking price feedback and capital support, resulting in a short economic cycle that is difficult to solidify. Prediction markets are reconstructing this logic by quantifying and pricing attention through price mechanisms. When users no longer just retweet or resonate but express judgments and beliefs with funds, attention begins to acquire economic depth. Football.Fun is transforming entertainment participation into a prediction pricing mechanism, where users are both participants and price setters, and the value of events is determined by the betting and verification process rather than narrative popularity.
Thus, prediction markets may become the core carrier of capitalizing attention resources, providing tradable price coordinates for collective interests in an era of information surplus and traffic scarcity. A deeper change lies in the migration of liquidity; unlike the meme ecosystem that relies on narratives and emotional fluctuations, the liquidity of prediction markets stems from event density and participation depth, possessing sustainability and compounding characteristics. Users are transitioning from speculators to evaluators, driving liquidity from emotional to functional migration. In the long run, the scale and activity of prediction markets are expected to surpass the meme ecosystem, becoming a core hub linking attention, cognition, and capital, and bringing a higher level of social pricing and cognitive aggregation to Web3 through a mechanism of tradable beliefs.
## Industry Landscape and Future Positioning
6.1. Challenges and Constraints
6.1.1. Regulatory Game
The biggest external variable for prediction markets currently still comes from regulatory uncertainty. The U.S. serves as a policy barometer in this field, but there are still conflicts between federal and state regulations. The CFTC views "event contracts" as regulated derivatives, while some state governments restrict them based on gambling laws, requiring platforms to deploy differentiated compliance modules in different jurisdictions, significantly increasing compliance costs. The regulatory dispute between Kalshi and the CFTC in 2025, while promoting discussions on the legalization of "event derivatives," has also exposed the gray areas of compliance implementation. In Europe and Asia, most countries are still in a "regulatory sandbox" or observation phase, and have not yet established a unified legal classification system. This institutional fragmentation leads to restricted capital flows, institutional funds being on the sidelines, and unstable compliance expectations, becoming the main obstacle to the globalization of prediction markets.
6.1.2. Liquidity Dilemma
Although prediction markets have formed a considerable scale, their depth of financialization is still constrained by liquidity design and funding source structure. Currently, mainstream platforms' trading is still concentrated on short-cycle, emotion-driven events (such as elections, sports, and trending topics), while long-term markets (technology, climate, macro trends) suffer from thin liquidity and limited price discovery functions. Additionally, market funding is still primarily from retail and crypto-native users, with institutional funds not yet systematically involved, leading to concentrated liquidity and increased risks of price manipulation. While oracle frameworks like Chainlink have improved the automatic settlement capabilities of objective events, some subjective events (such as policy decisions) still require manual arbitration. This lack of unified standards and transparent settlement mechanisms makes it difficult for prediction markets to be absorbed by mainstream financial systems.
6.1.3. User Mindset
The ideal participants in prediction markets should be information traders with rational judgment and risk assessment capabilities, but in reality, user behavior still tends toward short-term speculation and emotional betting. Moreover, with the integration of social dissemination and financialization, market volatility is further amplified. When there is a lack of sustained incentives, user retention and long-term participation decrease. For prediction markets to achieve a true "collective intelligence layer," they must establish more stable incentive and governance structures to balance short-term popularity with long-term value accumulation. In the absence of institutionalized returns and social reputation binding, their user base will long oscillate between speculation and entertainment, making it difficult to support the core logic of information financialization.
6.1.4. Information Asymmetry
The core logic of prediction markets lies in achieving probability pricing through collective intelligence, but in reality, information asymmetry and technical arbitrage are becoming significant sources of systemic risk. Some traders use technical means to capture key information in advance, causing prices to deviate from the true collective consensus. For example, in the recent Nobel Peace Prize event, abnormal trading and severe fluctuations in odds occurred on Polymarket hours before the results were announced, raising suspicions that someone had prior knowledge of the results. Subsequent investigations revealed that the information did not come from an internal leak but from technical monitoring of the Nobel official website. The website, based on a WordPress structure, had uploaded the winners' portrait files before the official announcement, which were indexed on publicly accessible websites. A developer used a script to continuously monitor the website's directory updates and inferred the winners' identities from the file names, placing bets on the prediction market in advance. This incident illustrates that prediction market prices are highly susceptible to the influence of information disclosure mechanisms and data transparency. When information can be extracted in advance through technical means, market pricing will be dominated by those with information advantages, undermining its credibility as a reflection mechanism of collective intelligence. Such cases indicate that prediction markets also face structural distortions caused by information gaps at the technical level, with prices reflecting the grasp of data channels by a few rather than public judgment.
6.2. Future and Growth Landscape
The optimal form of prediction markets may be to become a pricing layer for social expectations, an aggregator of market signals, and a probability engine for institutional decision-making. In the long run, its positioning may no longer be limited to single event betting or social sentiment games, but rather develop towards capital efficiency, trading density, and real-world impact.
6.2.1. Market Density
The vitality of prediction markets stems from the density of event supply and the sustainability of trading frequency. Currently, mainstream markets are still highly concentrated on political and election themes, with event structures exhibiting a dependence on hot topics, making it difficult to support a stable trading ecosystem. To achieve true scalable growth, prediction markets must build a multidimensional, continuous, and automatically generated event network covering technology innovation, macroeconomics, cultural entertainment, climate risks, and more. Emerging protocols like Limitless have attempted to create market structures with shorter settlement cycles, bringing trading behavior closer to the high-frequency iterations of derivatives. In the future, as open APIs, community self-themed creation, and AI-driven event generation gradually mature, prediction markets may evolve into continuously generated event flow exchanges, achieving around-the-clock liquidity cycles and user growth flywheels.
6.2.2. Leverage Innovation
Currently, the revenue structure of prediction markets remains linear and closed; binary markets under non-leverage conditions struggle to meet the return expectations of professional traders and cannot fully stimulate capital efficiency. Some protocols have begun to experiment with introducing leveraged predictions and options models, allowing users to amplify positions to increase marginal returns and rebalance risk and reward. However, constrained by the lack of standardized margin mechanisms and cross-market clearing systems, these products remain in the prototype stage. In the future, institutionalizing leverage mechanisms may become the primary breakthrough for the financialization of prediction markets. Through on-chain margin accounts, dynamic risk parameters, and multi-asset collateral systems, platforms can release higher capital utilization rates under controlled risk conditions. Additionally, if leveraged products are combined with governance incentives and reputation scoring systems, they will effectively enhance user retention and stickiness. As this structure matures, prediction markets will approach the mechanism level of derivatives markets, establishing the foundational conditions to attract institutional and professional liquidity providers.
6.2.3. Information Valorization
The fundamental value of prediction markets lies not in the predictions themselves but in whether the prediction results can become effective signals in the real world. When market prices are adopted by corporate strategies, policy-making, or financial models, prediction markets complete the transition from a layer of social entertainment to a layer of information pricing. In the future, the value of prediction markets may be reflected in the feedback effects of their results on the real world. When market prices are referenced by corporate strategies, policy assessments, or financial decisions, prediction markets will no longer be mere speculative platforms but become hubs linking social cognition and economic behavior. The essence of this process is to transform market probabilities into action signals. Prices will no longer just be the result of transactions but will embody the flow of information and decision-making efficiency. With the continuous improvement of oracle networks, reputation layers, and settlement standards, prediction markets will possess higher data reliability and institutional adoptability. In the long run, the development direction of prediction markets is not to expand the number of events but to enhance the signal value and real-world impact of results. When decision-makers at all levels of society, whether institutional investors, policymakers, or decentralized communities, can extract expectations, calibrate judgments, and optimize resource allocation from market prices, prediction markets will truly become the probability hub of the information economy, playing a role in pricing, coordination, and cognitive aggregation on a global scale.
## Reference Links
https://kalshi-public-docs.s3.amazonaws.com/kalshi-source-agency-trading-prohibitions.pdf
https://docs.polymarket.com/polymarket-learn/get-started/what-is-polymarket
https://limitlesslabs.notion.site/#19304e33c4b9808498d9ea69e68a0cb4
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
