Polymarket deployed canary markets on October 5, 2026, running its first production trades on Protocol V2 — a full smart-contract rebuild that replaces the Gnosis Conditional Tokens Framework the platform has used since 2019. The switchover for all new markets is tentatively scheduled for Novembe...
"Today marks the first complete rebuild of Polymarket's underlying smart-contract infrastructure since the protocol's inception." — Rajath Alex, Head of Protocol, Polymarket
Polymarket deployed canary markets on October 5, 2026, running its first production trades on Protocol V2 — a full smart-contract rebuild that replaces the Gnosis Conditional Tokens Framework the platform has used since 2019. The switchover for all new markets is tentatively scheduled for November 2, 2026.
The rebuild consolidates what had become a layered stack of adapters, wrapped tokens, and separate exchanges into a single ERC-1155 position contract, one collateral token (pUSD), one exchange per market type, and one router. Six audit firms reviewed the code, Certora formally verified it, and Polymarket posted a $5 million bug bounty on Cantina's platform.
The stakes are material. Polymarket has processed $72 billion in contracts traded in 2026 through September. The platform raised $1 billion in September 2026 at a $21 billion valuation, with investors including Donald Trump Jr.'s 1789 Capital, Intercontinental Exchange (NYSE's parent), and Peter Thiel's Founders Fund. Combined monthly prediction market volume across Polymarket and Kalshi reached $50 billion by July 2026, up from under $5 billion in September 2025.
Polymarket's V1 infrastructure was built atop the Gnosis Conditional Tokens Framework, a general-purpose smart contract system released in 2019. The framework was not designed for prediction market scale. Over time, Polymarket accumulated layers: separate adapters for pUSD collateral, wrapped-token contracts for negative-risk markets, multiple exchange contracts, and individual UMA oracle deployments per market type.
The Gnosis framework had structural limitations. Position tokens created through different splitting paths were incompatible — users betting on the same outcome could hold tokens that could not be merged if they were generated in a different order. This path-dependence problem fragmented liquidity and complicated market making.
V2 replaces the entire stack. According to Polymarket's protocol lead Rajath Alex, the rebuild goes "from the position token up," eliminating the accumulated technical debt.
Protocol V2 introduces a single ERC-1155 contract for all position tokens across every market type. Each position ID now encodes three pieces of information internally: market type, market identifier, and outcome. This eliminates the need for separate contract lookups or adapter layers.
The architecture consolidates to four components:
V2 launches with four market modules:
The modular design allows additional market types to be added without modifying the core position or collateral contracts.
V1 deployed individual UMA oracle instances for each market type. V2 introduces an OracleAggregator — a single contract that routes resolution requests to pluggable oracle modules.
At launch, the OracleAggregator supports UMA and Chainlink. The modular interface allows additional oracle providers to be integrated without redeploying the core contracts. According to Polymarket, this "widens the design space for how outcomes are settled."
The shift matters for market reliability. UMA's Optimistic Oracle, which Polymarket has relied on for resolution since inception, has faced periodic disputes and resolution delays. A multi-oracle architecture provides redundancy and allows market creators to select resolution mechanisms appropriate to the event type.
Under V1, Polymarket used a mixed collateral approach that had grown complex over successive product additions. V2 standardizes on pUSD as the sole collateral token. pUSD is redeemable 1:1 for Circle's USDC.
The standardization simplifies accounting, reduces the number of token approvals required for trading, and consolidates liquidity into a single collateral pool. For market makers — who accounted for 35.2% of trade volume by dollar value in Q1 2026, according to TRM Labs data — the single-collateral model reduces operational complexity.
V2 includes native infrastructure for bridging positions, collateral, and resolution data across chains. The bridging capability is built into the protocol architecture as a first-class feature rather than bolted on after the fact.
Polymarket has not activated cross-chain functionality at launch. The platform currently operates exclusively on Polygon. The inclusion of bridging infrastructure signals a planned multi-chain deployment, though no timeline has been disclosed.
Six independent firms audited the V2 codebase:
Certora's formal verification goes beyond standard auditing by mathematically proving that contract behaviors match their specifications under all possible inputs. The $5 million bug bounty program, hosted on Cantina's platform, covers critical vulnerabilities in the production deployment.
The canary period — small production markets running on V2 from October 5 through October 30 — provides a live testing environment before full switchover. V1 markets continue to operate and settle independently of V2 deployment.
Polymarket also released Data API V2, a Rust-based service running on its in-house indexer, alongside a four-week integration window and weekly office hours for developers and market makers transitioning to the new system.
The prediction market sector has scaled rapidly. Combined monthly volume across Polymarket and Kalshi rose from under $5 billion in September 2025 to more than $50 billion by July 2026, according to TRM Labs. August volume eased to $45.33 billion.
Polymarket-specific figures through September 2026:
Category distribution has shifted. Sports now accounts for approximately 39-40% of total volume — the 2026 FIFA World Cup alone generated $2.5 billion in wagered volume in its first 11 days. Politics accounts for 32% and crypto for 20%, according to Pew Research.
User segmentation data from TRM Labs (January–March 2026) shows: | Tier | Volume Share | |------|-------------| | Mid-frequency traders (11–1,000 fills) | $869M (44.7%) | | Market makers (10,000+ fills) | $774M (35.2%) | | Casual users (2–10 fills) | $3.5M (<0.2%) |
The median trade size for experienced users is $12, compared to $30 for newcomers — a pattern consistent with high-frequency participants placing smaller, more frequent trades.
Polymarket's market position has shifted since 2024. Kalshi, a CFTC-regulated designated contract market, now accounts for approximately 82% of monthly prediction market turnover, a reversal from late 2024 when Polymarket held over 90%.
For the week of September 28 to October 4, 2026, Kalshi volume reached $18.37 billion versus Polymarket US at $2.71 billion. Kalshi's gains have been driven primarily by sports contracts, where its regulated US status provides a compliance advantage over Polymarket's crypto-native model.
Polymarket retains dominance in politics and geopolitical markets. The platform's crypto-native infrastructure — on-chain settlement, pseudonymous trading, global access — attracts a different user base than Kalshi's KYC-heavy model.
Polymarket's $21 billion valuation (September 2026, led by 1789 Capital with $300 million) and $600 million earlier investment from Intercontinental Exchange reflect investor confidence despite the market share shift. ICE's involvement is notable: the NYSE's parent company backing a crypto-native prediction market suggests institutional conviction in the category's long-term economics.
Polymarket acquired QCEX LLC — a CFTC-licensed Designated Contract Market and Derivatives Clearing Organization — in July 2025 for $112 million. This acquisition provided a regulated pathway to offer contracts to US users.
Prediction markets are legal at the federal level in all 50 states under CFTC authority. However, state-level challenges persist. Eleven states introduced prediction market legislation in 2026, ranging from outright bans to taxation frameworks. Minnesota signed a ban in May 2026, but a federal judge issued a preliminary injunction finding it likely preempted by federal law. Rhode Island's attorney general sued both Kalshi and Polymarket; the CFTC countersued to assert federal jurisdiction.
The CFTC issued an Advanced Notice of Proposed Rulemaking in March 2026 soliciting public input on formal prediction market rules. Additionally, the CFTC's October 5 Regulation CTX proposal — focused on leveraged crypto trading — lists both Polymarket and Kalshi among DCM-registered exchanges positioned to participate in the new crypto asset market framework.
Protocol V2 is an infrastructure bet. Polymarket is replacing its core smart contracts while processing billions in monthly volume, running canary markets in production rather than waiting for a clean migration window. The six-audit, formal-verification, and $5 million bounty approach reflects the capital at risk in the system.
The competitive dynamics are worth watching. Kalshi's regulated exchange model has taken the volume lead, particularly in sports. Polymarket's crypto-native model — on-chain settlement, global access, pseudonymous trading — serves a different market. Whether V2's modular architecture and cross-chain infrastructure can recapture volume share depends on execution and, increasingly, on how state and federal regulators draw the boundaries of the prediction market category.
The November 2 switchover will be a live test at scale. Existing V1 markets will continue to settle independently, but every new market created after the cutover will run on the new architecture. For a platform valued at $21 billion, the margin for error is narrow.