Limitless Exchange

Summary

Limitless Exchange is a prediction market on Base, the Coinbase-incubated Ethereum L2. Founded in late 2023 by CJ Hetherington (previously co-founder of the web3 metaverse project Atlantis World) with co-founders reported as Roman Mogylnyi, Dima Horshkov, and Rev Miller, it launched publicly in 2024 with a mobile-first focus on very short-duration markets — hourly and daily questions about crypto price action ("Will BTC be above $100k this week?") — rather than the long-dated political markets that defined Polymarket. It later generalized into a full Polymarket-style venue: fully collateralized YES/NO outcome shares traded on a central limit order book (CLOB), with sports, esports, finance, politics, and pop-culture categories.

The project raised a reported ~$3M pre-seed led by 1confirmation, a $4M strategic round in July 2025 that brought in Arthur Hayes (via Maelstrom) as investor and advisor, and a $10M seed round announced October 20, 2025 led by 1confirmation with Coinbase Ventures, DCG, Collider, and others. Its LMTS token generated on October 22, 2025 (1B supply, ~24.37% to ecosystem incentives/airdrops) after seasons of points farming. By 2025 it was the largest prediction market on Base with roughly $500M cumulative volume; by 2026 the team and press cited around $2B in monthly volume and "$3B+ traded," placing it behind Kalshi and Polymarket in the category. It remains live and growing; outcome is ongoing.

Design (Mechanism)

• Fully collateralized binary shares. Every YES + NO pair is backed by $1.00 of USDC held in Gnosis's Conditional Tokens Framework (CTF) — the same primitive Polymarket uses. Shares trade between $0.01 and $0.99; winning shares redeem for $1.00, losers go to zero. Minting/splitting, merging, and redemption all happen against a shared CTF contract on Base (0xC9c9...6e18).
• CLOB execution, forked from Polymarket. Limitless openly forked the Polymarket CTF Exchange contracts (the GitHub repo is literally titled "Polymarket CTF Exchange for Limitless"). Orders are signed off-chain with EIP-712 against the venue's exchange contract as verifyingContract and settled on-chain — a hybrid off-chain-orderbook/on-chain-settlement model. Multiple exchange versions (V1–V3) plus parallel "NegRisk" exchanges support multi-outcome (categorical) markets with share conversion between outcomes.
• Centralized multisig resolution. Unlike Polymarket's UMA optimistic-oracle path, Limitless uses a GnosisSafe multisig as the oracle: the team's safe calls prepareCondition and reports outcomes directly on the CTF. This enables the "instant resolution" UX that hourly price markets require, at the cost of trust in the operator.
• Short-duration market focus. The signature product is nonstop hourly/daily crypto price markets that resolve quickly — closer in spirit to binary options than to election forecasting — marketed as leverage-like upside "without liquidation risk."
• Incentive flywheel. Limitless Points seasons, order-book LP rewards for quoting inside the spread, and a 24.37% ecosystem allocation of LMTS (Season 1 airdrop unlocked fully at TGE) drove volume farming; LMTS staking ("earn yield on LMTS") was added post-TGE. In June 2026 it launched its first permissionless market type using a deliberately templated, finance-scoped approach to market creation rather than fully open curation.

Outcome

Ongoing and, so far, a category winner on its chain. Limitless grew from ~$9.7M volume in its early daily-markets phase to ~$500M cumulative by late 2025, then to reported ~$2B/month during the 2025–26 prediction-market boom, making it the largest prediction market on Base. It raised ~$17M+ across rounds, shipped its LMTS token in October 2025, and attracted Arthur Hayes as advisor. Founder CJ Hetherington became a prominent public voice in the sector — notably in October 2025 alleging (his claim, disputed by the exchange) that Binance asks projects to reserve ~9% of token supply for listing-related programs. Long-run durability is untested: volumes are heavily incentive-driven (points seasons ran through at least May 2026), and LMTS faces ongoing unlock overhang (e.g., ~85M tokens on April 22, 2026).

Why it worked

• Fork the winner, differentiate on tempo. Reusing Polymarket's audited CTF/CLOB stack meant Limitless spent zero innovation budget on settlement plumbing and all of it on product: hourly/daily crypto price markets with instant resolution — a niche Polymarket underserved and one that matches crypto traders' dopamine cycle.
• Right chain, right patron. Being Base-native gave it cheap fees, Coinbase-adjacent distribution, and a lane where it did not compete head-on with Polymarket (Polygon) or Kalshi (regulated US) — it became "the" prediction market of its ecosystem.
• Aggressive but legible incentives. Points seasons, LP rewards for tight quoting, and a large airdrop allocation bootstrapped both sides of the order book before the token existed.
• Timing. It scaled into the 2024–25 prediction-market supercycle (US election, Polymarket/Kalshi mainstreaming) with a differentiated product already live.

Limitations and criticisms

Resolution trust is centralized: a team multisig reporting outcomes is a step back from the optimistic-oracle decentralization Polymarket pioneered, making the operator a single point of failure for every market. Hourly price markets are economically close to binary options/gambling. Volume quality is uncertain — much of it was farmed against points seasons and LMTS emissions, and the token carries a heavy unlock schedule against a 25% team / 25% investor allocation. Finally, it remains a fast follower in a market where Polymarket and Kalshi command distribution and brand; its own CEO argues no platform will dominate the category, an assessment that also applies to Limitless itself.

Lessons

• In mechanism-heavy verticals, forking settled infrastructure (CTF + CLOB) and competing on market selection, chain, and UX is a viable and capital-efficient wedge.
• Resolution latency is a product axis: decentralized oracles (UMA) trade speed for trust; a multisig oracle trades trust for the instant settlement that short-duration markets require. Pick per market type, and be honest about the trust model.
• Short-duration, high-frequency markets generate volume far faster than long-dated event markets — but that volume is closer to options flow than to forecasting, with the retention profile that implies.
• Points-to-token flywheels can bootstrap an order book, but cumulative-volume milestones earned under emissions are weak evidence of organic demand; the test is retention after seasons end.
• Ecosystem-native positioning (the biggest X on chain Y) is a real moat when the L2's patron (Coinbase/Base) is motivated to showcase you.

Redesign (EDITORIAL — hypothesis, not fact)

This section is editorial hypothesis, not reporting. The most fragile element is the GnosisSafe oracle. A redesign could keep instant resolution for price markets by resolving them from on-chain/oracle price feeds (e.g., Chainlink or a TWAP attested at expiry) with the multisig only as a fallback, while routing subjective markets through an optimistic oracle with a bonded dispute window — tiering trust by market type instead of applying maximum trust everywhere. Second, incentive design could shift from volume-farmed points (which reward wash-like churn) toward maker-quality rewards: pay for time-at-best-quote, depth within the spread, and realized fill quality, decaying rewards for self-crossing flow. Third, given the binary-options character of hourly markets, a sustainable version likely needs proactive market-integrity tooling (position limits near resolution, oracle-price freeze windows). Finally, LMTS would be healthier with utility tied to resolution security — e.g., staked LMTS backing a dispute/insurance pool — rather than generic staking yield layered on an emissions token.