FanDuel Predicts is a CFTC-regulated prediction market platform built as a joint venture between FanDuel Group and CME Group. Launched December 22, 2025, it allows users in all 50 US states to trade Yes/No event contracts on sports, finance, politics, and economics — operating under federal derivatives law rather than state gaming regulation. This guide covers the platform’s architecture, fee structure, market categories, regulatory framework, developer limitations, and how it compares to Kalshi, Polymarket, and DraftKings Predictions.
What FanDuel Predicts Is
FanDuel Predicts is a standalone prediction market app, separate from the FanDuel Sportsbook. The core difference: instead of betting against a bookmaker at house-set odds, users trade binary event contracts with each other on a peer-to-peer exchange. Contracts are priced between $0.01 and $0.99, with the price reflecting the market’s implied probability of an outcome. Correct predictions settle at $1.00; incorrect ones settle at $0.00.
The platform is the product of a partnership between FanDuel Group (a subsidiary of Flutter Entertainment, NYSE: FLUT) and CME Group (NASDAQ: CME), the world’s largest derivatives marketplace. CME Group lists the event contracts on its exchanges and handles clearing. FanDuel provides the consumer-facing app, customer acquisition, and KYC infrastructure. CME receives 50% of gross revenue from the venture, while FanDuel covers all operational costs.
Within the Agent Betting Stack, FanDuel Predicts sits primarily at Layer 3 (Trading) — it is an exchange for executing positions on event outcomes. However, because it lacks a public API, it currently has no Layer 4 (Intelligence) integration path for AI agents. This is a meaningful limitation compared to platforms like Kalshi and Polymarket, which provide programmatic access for automated trading systems.
Regulatory Structure
FanDuel Predicts operates under a fundamentally different legal framework than traditional sportsbooks. The contracts are classified as financial derivatives regulated by the Commodity Futures Trading Commission (CFTC), not as gambling products regulated by state gaming boards. This distinction is what allows FanDuel Predicts to operate in states where sports betting remains illegal.
Accounts are opened through FanDuel Prediction Markets LLC, which is a registered futures commission merchant (FCM) and a member of the National Futures Association (NFA). Users must complete full KYC verification, including submitting their birth date, Social Security number, home address, banking information, and a government-issued ID. The onboarding process is identical to opening a brokerage account, not a sportsbook account.
The platform’s state-by-state availability breaks down into two tiers. Non-sports markets — covering finance, economics, politics, crypto, and science — are available in all 50 states. Sports event contracts are available only in 18 states where online sports betting is not yet legal: Alabama, Alaska, California, Delaware, Florida, Georgia, Hawaii, Idaho, Minnesota, Nebraska, New Mexico, North Dakota, Oklahoma, Rhode Island, South Carolina, South Dakota, Texas, and Utah. FanDuel has stated publicly that sports contracts will be withdrawn from any state that subsequently legalizes online sports betting.
The legal landscape around event contracts remains unsettled. Multiple lawsuits are active challenging whether sports event contracts constitute illegal gambling under state law. Wall Street analysts expect this question to eventually reach the US Supreme Court. For context on the broader regulatory picture, see our guide on whether prediction markets are legal.
How Trading Works
The trading mechanics on FanDuel Predicts follow the standard prediction market model. Each market poses a binary question — for example, “Will the Cowboys win 10+ games?” or “Will CPI exceed 3.2% in the next report?” Users buy Yes or No contracts at the current market price, which functions as an implied probability.
| Concept | How It Works |
|---|---|
| Contract pricing | $0.01–$0.99 per contract, reflecting implied probability |
| Settlement | Correct side pays $1.00; incorrect pays $0.00 |
| Position exit | Sell contracts back into the market before settlement (subject to liquidity) |
| Order types | Market orders and limit orders are both supported |
| Payout timing | Pending payouts credited immediately; withdrawable in 1–2 business days |
The key difference from sportsbook betting: prices are set by other traders, not by a bookmaker. If you think the market is mispricing an outcome, you buy or sell contracts to express that view. This creates genuine price discovery rather than house-adjusted odds.
FanDuel Predicts supports both market orders (executed immediately at the best available price) and limit orders (executed only at a specified price). The ability to place limit orders is a notable advantage over some competitor platforms — though the platform does not expose a full order book to retail users, which limits visibility into market depth.
Market Categories
FanDuel Predicts organizes its markets into several categories. The depth and breadth of each category varies significantly.
Sports (18 states only): Contracts on NFL, NBA, NHL, MLB, and college sports outcomes. These include season-long futures (e.g., “Will the Cowboys win 10+ games?”), game-level outcomes, and threshold-based contracts. Sports markets are the primary draw for users in states without legal sportsbooks. For detailed analysis of FanDuel’s traditional sportsbook odds quality, see our FanDuel odds overview and FanDuel prop bets guide.
Finance and Economics: Contracts on S&P 500 and Nasdaq-100 benchmarks, oil and gas prices, gold, cryptocurrency prices, CPI prints, GDP reports, and Federal Reserve interest rate decisions. These are available nationwide and represent the platform’s most “financial” offering.
Politics and World Affairs: Contracts on election outcomes, policy decisions, and geopolitical events. This category saw heavy traffic during the 2026 midterm cycle.
Science and Technology: Contracts on space exploration milestones, climate records, and tech industry events.
Culture and Mentions: A unique category where users can trade on whether public figures will use certain words during speeches or press briefings, entertainment outcomes, and cultural moments.
Compared to Kalshi, which lists proprietary contracts across dozens of subcategories, FanDuel Predicts’ catalog is narrower. Compared to Polymarket, which covers global events and crypto-native topics with deep liquidity, FanDuel is more conservative in scope. The trade-off is regulatory clarity and mainstream accessibility.
Fee Structure
FanDuel Predicts’ fee structure has two layers: the platform fee and the CME exchange fee.
| Fee Component | Amount | When Charged |
|---|---|---|
| FanDuel platform fee | ~$0.02 per $1 of potential payout (2%) | Deducted upfront at trade entry |
| CME Globex exchange fee | $0.01 per contract, per side | On open and close |
| Cash settlement fee | $0.00 | At settlement |
| Withdrawal fee | None | N/A |
For a practical example: if you buy a Yes contract at $0.30 (potential payout of $0.70 per contract), the platform fee is approximately $0.014 (2% of $0.70). The CME exchange fee adds $0.01 per side, so $0.02 round-trip. Total cost on a $0.30 contract held to settlement: roughly $0.034 in fees.
The fee opacity is a noted concern for quantitative traders. FanDuel does not publish a granular, formula-based fee schedule the way Kalshi does. Active traders should benchmark their fills against other platforms offering the same CME contracts to calculate their true cost of execution. For thin-edge strategies, this opacity can be a dealbreaker — if your model identifies a $0.03 mispricing and fees consume $0.02 of that edge, the remaining expected value may not justify the capital at risk.
On March 2, 2026, FanDuel eliminated credit card funding. Accounts can be funded via bank transfer, debit card, PayPal, or Apple Pay.
Developer and AI Agent Access
This is where FanDuel Predicts falls short for the AgentBets audience. Unlike Kalshi, which provides a full REST API with WebSocket streaming, or Polymarket, which offers CLOB, Gamma, and Data APIs for programmatic trading, FanDuel Predicts does not have a public developer API.
The contracts are listed on CME Group exchanges and technically trade on the CME Globex electronic trading platform. CME does offer API access to Globex for institutional participants — but this is designed for registered FCMs and large-scale market makers, not for retail developers or hobbyist bot builders. There is no documented path for an individual developer to programmatically access FanDuel Predicts order books through CME’s infrastructure.
This means AI agents built on frameworks like CrewAI or OpenClaw cannot currently execute trades on FanDuel Predicts. The platform is app-only. Where it fits in the broader ecosystem is as a data signal: FanDuel Predicts prices reflect the probability estimates of a large, mainstream retail audience (FanDuel has approximately 17 million registered users). These prices can serve as inputs to an agent’s intelligence layer — but the agent cannot act on them directly within FanDuel Predicts.
For builders looking to construct automated trading pipelines, Kalshi and Polymarket remain the only viable CFTC-regulated and crypto-native options respectively. See our prediction market API reference for a complete comparison.
FanDuel AceAI: What It Is and Isn’t
FanDuel’s AceAI is a generative AI chatbot that launched in March 2025, built into the FanDuel Sportsbook app. It allows users to research player statistics, surface performance trends, and construct Same Game Parlays through natural language queries. For example, a user can ask AceAI to build a parlay with specific player prop thresholds, and the system returns odds within seconds.
AceAI is relevant to the prediction market discussion because it demonstrates FanDuel’s investment in AI-assisted trading tools. However, as of April 2026, AceAI is a Sportsbook feature — it has not been integrated into FanDuel Predicts. There is no AI assistant within the Predicts app for analyzing event contract markets, researching economic indicators, or suggesting trades.
AceAI’s initial rollout was limited to approximately 1% of FanDuel customers and covered only NFL and NBA markets. It includes responsible gambling features that detect potentially problematic betting behavior and direct users to support resources. The technology was developed under Jon Sadow, FanDuel’s VP of Product Transformation and Innovation.
Whether FanDuel brings AceAI-style functionality to Predicts is an open question. The convergence of AI tools and prediction markets is something we track closely in our coverage of the sportsbook–prediction market convergence.
FanDuel Predicts vs. Kalshi vs. Polymarket vs. DraftKings Predictions
The US prediction market landscape in 2026 has four major players. Here is how they compare across the dimensions that matter most.
| Feature | FanDuel Predicts | Kalshi | Polymarket | DraftKings Predictions |
|---|---|---|---|---|
| Regulator | CFTC (via CME) | CFTC (direct) | CFTC (via QCEX) | CFTC (via CME) |
| Exchange operator | CME Group | Kalshi (self-operated) | Polymarket (self-operated) | CME Group |
| US availability | All 50 states | 43+ states | Expanding (waitlist) | 38+ states |
| Sports contracts | 18 states | Limited states | Yes (varies) | Varies by state |
| Public API | No | Yes (REST + WebSocket) | Yes (CLOB + Gamma + Data) | No |
| Settlement currency | USD | USD | USDC (crypto) | USD |
| Fee model | ~2% of potential payout | Tiered by price | ~0.75% (sports), varies | ~2% of potential payout |
| Min deposit | $1 | $10 | None (requires USDC) | Varies |
| KYC required | Yes (SSN, ID) | Yes (SSN, ID) | Yes (for US relaunch) | Yes (SSN, ID) |
| AI agent compatible | No | Yes | Yes | No |
| Parent company | Flutter Entertainment | Independent | Independent | DraftKings |
For a deeper comparison of DraftKings’ prediction market offering, see our Kalshi vs. DraftKings Predictions analysis and DraftKings Predictions Guide.
Who FanDuel Predicts is best for: Retail users in non-sports-betting states who want a familiar, trusted brand to trade sports event contracts. Users who already have FanDuel accounts benefit from seamless fund transfers between Sportsbook and Predicts. The platform’s mainstream UX makes it the easiest on-ramp for users who have never traded prediction markets before.
Who should look elsewhere: Developers building AI agents or trading bots (no API). Quantitative traders who need transparent fee schedules and full order book visibility (Kalshi). Crypto-native traders who want the deepest global liquidity and lowest fees (Polymarket). Traders who want broad market diversity beyond sports and macro finance (Kalshi).
The Strategic Play: Why FanDuel Built This
FanDuel Predicts is not primarily a prediction market product — it is a customer acquisition strategy. The platform gives FanDuel a presence in California, Texas, and Florida, the three largest US states by population, none of which have legalized online sports betting. Combined, these three states represent over 90 million residents that FanDuel’s Sportsbook cannot reach.
Flutter Entertainment, FanDuel’s parent company, is investing heavily: the company projects $200–300 million in incremental EBITDA costs during 2026 to build out Predicts. That level of investment signals that Flutter views prediction markets as a long-term customer acquisition channel, not a standalone profit center — at least not yet.
The strategic logic maps to a specific scenario: if any of these states legalize online sports betting, FanDuel already has a verified, depositing customer base ready to migrate to the Sportsbook product. Even if legalization does not happen, FanDuel captures trading activity on finance, politics, and culture markets that generate revenue through fees.
From an odds and vig analysis perspective, FanDuel Predicts operates on a fundamentally different margin model than the Sportsbook. Traditional sportsbook vig averages 5–8% across markets. Prediction market fees of 2% plus exchange costs are structurally lower, which means the revenue model depends on volume rather than margin. FanDuel’s existing user base of 17 million registered customers gives it a significant liquidity advantage if it can drive adoption.
FanDuel Odds Pages on AgentBets
For users who also bet on FanDuel’s traditional sportsbook, AgentBets provides comprehensive odds and vig analysis. These pages are updated daily and cover every sport FanDuel offers.
Core pages:
- FanDuel Odds Today — Overall vig analysis and sport-by-sport breakdown
- FanDuel Sportsbook Review — Complete review covering odds quality, state availability, and promotions
- FanDuel Prop Bets Guide — Same Game Parlay mechanics and prop market analysis
- Odds Boost Comparison — FanDuel boost analysis alongside DraftKings, BetMGM, and others
Head-to-head odds comparisons:
- DraftKings vs. FanDuel — The rivalry matchup
- BetMGM vs. FanDuel — Third-largest regulated book vs. FanDuel
- Bovada vs. FanDuel — Offshore vs. regulated comparison
- FanDuel vs. Pinnacle — FanDuel against the sharpest book
- BetOnline vs. FanDuel — Another offshore alternative
For a three-way comparison covering vig, odds boosts, and plus-EV strategy across the big three regulated books, see DraftKings vs. FanDuel vs. BetMGM Odds Comparison.
What FanDuel Predicts Means for AI Agents
The emergence of FanDuel Predicts — alongside DraftKings Predictions, Fanatics Markets, and established platforms like Kalshi — confirms that event contract trading is no longer a niche crypto experiment. It is a regulated, institutional-grade product category backed by CME Group, Flutter Entertainment, and DraftKings.
For AI agent builders, the implications are straightforward. The opportunity is real — prediction markets are growing rapidly, with industry analysts projecting annual trading volume could reach $1 trillion by the end of the decade. But the execution path remains constrained by API availability. FanDuel Predicts and DraftKings Predictions are closed ecosystems. Kalshi and Polymarket are the only platforms where agents can actually trade.
The most productive approach for now: use FanDuel Predicts prices as one signal among many in an agent’s intelligence layer, while routing actual execution through Kalshi or Polymarket APIs. As the market matures, we expect API access to expand — but the sportsbook operators have no incentive to open their platforms to algorithmic traders who would extract value from retail flow.
For a complete framework on how AI agents interact with prediction markets and sportsbooks across all four layers of the stack, start with our Agent Betting Stack architecture guide.