Virtuals Protocol price

Written by ChandlerZ, Foresight News If human society has been curious and betting on the future since ancient times, then crypto-native prediction markets are transforming this ancient demand into public goods that can be priced, liquidated, and reused. The democratization of information has been done by the Internet in the past decade; In the Web3 and crypto fields, values and beliefs are also being tokenized and priced, forming a more verifiable and incentive-compatible democratization of value. The addition of AI has shifted the boundaries of forecasting from simple price feeding to more complex judgments and rulings, giving forecasting the meaning of infrastructure, eliminating speculative understanding, and the prediction market is the underlying information cornerstone of governance, hedging and resource allocation. In November 2025, Google began to integrate the market probability of Polymarket and Kalshi into the Google Finance scenario, marking that the prediction data is entering the public network layer of the 100 million user entrance, which is not only an endorsement of the industry, but also a signal of incremental demand. Why prediction markets are a must for Web3 The essence of the prediction market is to aggregate the dark knowledge scattered in the minds of individuals into common probabilities with prices. This idea can be traced back to Robin Hanson's Futarchy, where prediction markets are set as the primary mechanism for information aggregation in a system where value targets are determined by voting and factual judgments are priced by the market. Academic research also shows that prediction markets are better at portraying event results than simple polls in many scenarios, especially in dynamic updates and incentive constraints. If you pull the perspective from theoretical reasoning back to the real market, you will find that this mechanism of price aggregation cognition is being voted on by funds and users with their feet in 2024-2025. Prediction platforms represented by Polymarket and Kalshi have repeatedly approached or even exceeded $100 million in daily trading volume, and the cumulative turnover has jumped to tens of billions of dollars, marking that the prediction market is moving from a niche experiment to a full-scale explosion. According to the data, Polymarket reached an all-time high of 477,850 monthly active traders in October, surpassing the previous record of 462,600 set in January. Its monthly trading volume also rebounded to a record $3.02 billion last month, after being around $1 billion or less from February to August. The number of new markets opened on the platform reached 38,270 in October, almost three times as many as in August. Polymarket's trading volume, active traders, and new market openings all hit record highs in October. Kalshi even surpassed Polymarket in October, reaching $4.4 billion. In addition, after the US regulation shifted to mergers and acquisitions with regulated entities, the path for compliance to return to the United States is becoming clearer. Together, this series of events shows that the information derivatives market with prediction as the core has real, strong and recognized demand by mainstream entrances. From the perspective of application spillover, the prediction market can be regarded as a general risk hedging and governance module, enterprises can hedge business risks on the probability of policy implementation, DAOs can use conditional markets to bind proposals and KPIs, and media and platforms can use probabilistic narratives as a new information display layer. The opening of information portals and prediction platforms such as Google and Perplexity is accelerating this era of probability-i.e. interface. The dilemma of investors under the prosperity of the track is available but not invested When a track enters an early breakout, ordinary investors usually ask two questions. One is whether the demand is real, and the other is how to share growth. We have already seen the answer to the former; The latter has long fallen into an embarrassing reality in the prediction track, and the head products are available and cannot be invested. Taking Polymarket as an example, its official once stated that the project does not have tokens and has not announced any airdrop or TGE plans. Although recently Polymarket CMO Matthew Modabber confirmed the POLY token and airdrop plans. Earlier in October, the company's founder, Shayne Coplan, had also revealed that it would launch the POLY token. But this still means, and it means, that for investors who did not participate deeply in Polymarket in the early days, the most generous and asymmetrical original dividend period has actually been consumed in advance. Unless you personally participate in every event market, it's hard to get track-level beta exposure aligned with long-term earnings. For investors who want to hold track growth in an exponential way, the target is extremely scarce. More broadly, regulated event contract platforms such as Kalshi also do not have crypto native tokens; Other prediction applications or tools on the chain are either not enough to play an industry index in terms of volume and network effects, or are more like single-function tools, making it difficult to carry track-level value attribution. As a result, demand blooms violently at the application layer, but there is a structural gap in the investment layer with no currency to invest in. From Pump.fun vs. Virtuals, look at Polymarket vs. DeAgent AI Looking back at the Meme track in 2024, one of the most representative phenomena is the emergence of Pump.fun, with an extremely low threshold and a standardized curve issuance mechanism, igniting the zero-to-one creation on the chain. In its early stage, the platform itself did not have a native token, and users could only share the prosperity by participating in individual stock games for each meme. Subsequently, the market emerged as a token carrier that can index this eco-level popularity, Virtuals (VIRTUAL). By binding key paths such as creation, trading, and LP matching within the ecosystem to the platform token, holding VIRTUAL is similar to holding the growth index of the entire Agent/Meme ecosystem, thereby inheriting the premium released by Pump.fun in terms of narrative and fundamentals. Pump.fun launched the platform token PUMP in mid-to-late 2025, but the timing is more later, and its value capture logic is misaligned with the previous ecological outbreak. Historical experience tells us that when the application layer explodes first and there is a lack of index assets, infrastructure projects that provide both products and tokens first often outperform the track average in value revaluation. Back to the shaping prediction market track, DeAgent AI is such an infrastructure-based role. DeAgentAI is an AI agent infrastructure covering the Sui, BSC, and BTC ecosystems, empowering AI agents to achieve trustless autonomous decision-making capabilities on-chain. It aims to address the three major challenges faced by AI in distributed environments: identity authentication, continuous assurance, and consensus mechanisms, and build a trustworthy AI agent ecosystem. DeAgent AI has built an underlying protocol centered on prediction markets and DeFi scenarios, with AI oracles and multi-agent execution networks as the core. One end connects real-world and on-chain data, standardizes complex judgments, rulings, and signal production into verifiable oracle outputs, and the other ends connect these outputs to trading, governance, and derivatives design through agent networks, thus becoming the information and value hub of the entire track. Because of this, this mirror is predicting a repeat of the market track today. Polymarket corresponds to the Pump.fun of the year (a leading product but long-term absence of investable tokens), while DeAgent AI (AIA) plays the role of a value container like Virtuals. It not only provides the key infrastructure modules (AI oracles and agent execution networks) that the prediction market lacks, but also provides the publicly tradable token AIA as an anchor for track indexation, allowing investors to indirectly share the medium- and long-term growth of the entire prediction track by holding AIA. How DeAgent AI is a value container for predicting the track The core of DeAgentAI's technical framework lies in solving the three fundamental challenges faced by decentralized AI agents running on-chain: continuity, identity, and consensus. Through the state system that combines hot memory and long-term memory, as well as on-chain state snapshots, agents will not be reset in multi-chain multi-tasking, and behaviors and decisions have a complete and traceable life cycle. Use on-chain unique identity + DID and hierarchical authorization mechanism to ensure that each agent's identity cannot be forged; Then, the Minimum Entropy Decision and the validator consensus are used to converge the chaotic output of the multiple models into a deterministic conclusion that can be settled. On this basis, the A2A protocol is responsible for standardized collaboration between agents, and the MPC execution layer ensures the privacy and security of sensitive operations, ultimately integrating identity, security, decision-making, and collaboration into a verifiable and scalable decentralized AI agent infrastructure. The two-wheeled landing of AlphaX and CorrAI At the application layer, AlphaX and CorrAI are the most intuitive realization of this infrastructure. AlphaX is the first AI model incubated by its community based on the DeAgentAI feedback training mechanism, employing a Transformer architecture, Mixture-of-Experts (MoE) technology, and a human feedback reinforcement learning (RHF) mechanism to focus on improving the accuracy of cryptocurrency price predictions. AlphaX predicts 2–72-hour crypto price trends, giving an accuracy rate of 72.3%, and achieving ROIs of +18.21% and +16.00% in live simulations in December 2024 and January 2025, with a win rate of around 90%, proving the considerable utility of AI predictions in real trading environments. CorrAI is more like a no-code Copilot for DeFi/quantitative users, helping users select strategy templates, adjust parameters, do backtesting, and issue on-chain instructions, stringing together visible signals and executing strategies into a closed loop, and also introducing more real funds and behaviors into DeAgent AI's agent network. On the ecological side, AlphaX has accumulated considerable users and interactions through activities and integrations on public chains such as Sui and BNB, and with multiple chains and multiple application scenarios, the overall network of DeAgent AI has formed a production relationship of hundreds of millions of on-chain interactions and tens of millions of users. From price feeds to AI oracles for subjective judgment Traditional oracles mainly handle objective values such as BTC/USD, relying on multi-node redundancy and data source aggregation to reach consensus. Once the question becomes subjective/indeterminate (e.g., "Is ETH more likely to rise or fall this weekend?"). The nodes call the large model separately, and the answers given are often inconsistent, and it is difficult to prove that a certain model is indeed called as agreed and this result is obtained, and security and trust begin to fail. DeAgent AI designed the DeAgentAI Oracle to address these subjective problems from the beginning. Users submit questions in the form of multiple-choice questions and pay service fees, and multiple AI agents in the network vote after independent judgment based on retrieval + reasoning. In this way, the originally divergent AI output is compressed into a deterministic result that can be settled, and believe it or not, a node is replaced with a set of public voting and recording processes that verify it, allowing AI to make judgments for the first time as a public service that can be repeatedly invoked on-chain, which is very suitable for scenarios such as prediction markets, governance rulings, and InfoFi. The component is currently being tested internally. In a specific case, DeAgent AI's Agents have been used to make judgments around real-world events. During the recent U.S. federal government shutdown, the team built a decision tree model at the end of October based on the market pricing of platforms such as Kalshi and Polymarket, combined with the duration of historical shutdowns, the structure of the bipartisan game, and key time nodes, and finally concluded that this round of shutdowns is most likely to be forced to end in the November 12-15 range (or a close range of November 13-20), rather than the infinite pull narrative common in market sentiment. During the same period, around the controversial topic of "whether Bitcoin has entered a bear market", DeAgent AI judged that the current stage is closer to the "early in-depth adjustment of the bear market" by integrating on-chain data, ETF capital flows, macro policy shifts, and technical indicator divergence, rather than an unfinished round of accelerated bull market. On the one hand, this kind of prediction and judgment around specific issues demonstrates the ability of DeAgent AI oracle to disassemble and integrate subjective and complex problems, and on the other hand, it also shows that its output can be directly transformed into signals that can be used to predict market and trading decisions, rather than just staying at the demonstration level. How AIA indexes track growth Returning to the investor's perspective, the value capture logic of AIA lies in the fact that it is not only a payment and settlement medium for the DeAgentAI Oracle and Agent networks, but also a staked asset and governance credential for nodes and validators. As more prediction applications, governance modules, and DeFi strategies connect to this network, the number of requests, call frequency, and security requirements will be transformed into actual demand for AIA, naturally tied to the usage of the entire track, rather than just relying on one-time narrative heat. More importantly, the value chain itself is closed-loop and deducable. Forecasting applications such as Polymarket need to rely on AI oracles to make complex judgments when expanding market categories and introducing more complex subjective questions. These calls will be directly reflected in the growing demand for AI oracle infrastructure such as DeAgent AI; As the usage of the Oracle/Agent network increases, the demand and value of AIA, the function token tied to it, will also rise as a payment, settlement, and staking asset. In other words, if you believe that the prediction market will continue to expand, it's hard not to believe that the demand for AI oracles will expand at the same time, and this will eventually be reflected in AIA's long-term pricing. In terms of asset attributes, AIA meets both the conditions of "function" and "investability". On the one hand, it corresponds to the AI oracle and agent infrastructure facing subjective problems, which directly connects with the core pain points of the prediction market. On the other hand, it is itself a token asset that can be allocated on the open market. In contrast, prediction platforms such as Kalshi and Polymarket have no native tokens to invest in, and although traditional price oracles have tokens, they serve an objective price feed track, which is not the same value chain as AI-led subjective oracles. In the subdivision of AI oracle + tradable tokens, AIA is currently one of the few or even the only targets that can meet both available and investable targets, so it has the opportunity to become the most direct indexing carrier for predicting the growth of the track. How should we participate in the prediction track? The current prediction track has obviously entered the stage where application stories are told in front of the stage and the value slowly sinks to the stage. Polymarket and Kalshi have proven the existence of the track with real transaction volume, and what can really be priced for a long time is likely to be the layer that supports the operation of these applications, that is, the AI oracle responsible for judgment and settlement, the agent network, and the functional tokens tied to it. As predictive applications try to carry more complex and subjective judgments, they are bound to have higher and more frequent call requirements for AI oracles. This demand eventually builds into continued use of infrastructure like DeAgent AI; The functional tokens closely tied to the payment, settlement, and staking of this infrastructure will also undertake corresponding value in this process. Then what we really need to think about next is no longer whether to participate in this track, but how and at what level to participate in this track. A relatively clear approach is that the application layer uses participation and the infrastructure layer uses positions. On the application layer, users can continue to use platforms such as Polymarket as tools to obtain alpha and use positions to bet on specific events. At the infrastructure layer, AIA is moderately configured to align AI oracles with the longer-term proposition of standard equipment in prediction markets. The former answered whether this game could make money, and the latter answered whether it was lifted up with the bottom when the track became bigger. Of course, AIA is only a factor in the portfolio, not an alternative to risk control itself. A safer way to open it is to treat it as part of the forecast track infrastructure index, and give this long-term logic a place and time within its own risk budget to allow the market to verify its judgment on this narrative.