On-chain perpetual futures exchanges processed $633.3 billion in 30-day volume as of October 3, 2026, according to DefiLlama data. That figure represents approximately 12.33% of the global perpetual futures market, up from 0.11% in January 2023 and 2% in January 2024. Hyperliquid, a custom Layer-...
"We will continue to invest in building the best possible infrastructure for decentralized trading." — Jeff Yan, Hyperliquid Labs Co-Founder
On-chain perpetual futures exchanges processed $633.3 billion in 30-day volume as of October 3, 2026, according to DefiLlama data. That figure represents approximately 12.33% of the global perpetual futures market, up from 0.11% in January 2023 and 2% in January 2024. Hyperliquid, a custom Layer-1 chain with a central limit order book, commands 30.9% of perp DEX weekly volume and 56.8% of on-chain open interest at $8.3–9.2 billion. The platform's cumulative protocol revenue surpassed $1.4 billion in October 2026, with $1.26 billion directed to open-market HYPE token buybacks.
The competitive landscape has fragmented. Aster (formerly Astherus, merged with APX Finance) holds the second position with approximately $40 billion in 30-day volume and $1.8 billion in open interest. Lighter, a zero-knowledge rollup exchange built by former Citadel engineers, sits third at roughly $35.7 billion in 30-day volume. EdgeX ($36.8 billion in 30-day volume) and Variational ($19.5 billion in seven-day volume as of early October) round out the top five. Despite this growth, centralized venues — led by Binance ($69 billion daily) and OKX — still process roughly 88% of all perpetual futures volume.
The structural shift from centralized to decentralized derivatives execution mirrors a pattern visible in spot markets during 2020–2022. Whether perp DEXs can sustain 12% market share — or compress back toward single digits as incentive programs expire — depends on whether their fee economics, latency, and liquidity depth can retain traders acquired through token subsidies.
The trajectory is unambiguous. On-chain perpetual DEX volume hit a record $1.2 trillion in October 2025, according to The Block. After a five-month decline through Q1 2026, 30-day volumes recovered to $633.3 billion by early October 2026. Monthly figures for September 2026 showed $593.3 billion in on-chain perp volume, per DefiLlama.
The growth rate in context:
| Period | Monthly Perp DEX Volume | DEX Share of Global Perps | |---|---|---| | January 2023 | ~$2B | 0.11% | | January 2024 | ~$82B | 2.0% | | October 2025 | $1.2T (record) | ~26% | | March 2026 | $340B (trough) | ~8% | | September 2026 | $593B | 12.33% |
The October 2025 peak coincided with aggressive token incentive programs across multiple platforms. Aster alone surged from $1 billion to $70 billion in daily volume within a single week during late September 2025, capturing roughly 70% of all perp DEX volume before activity normalized. That episode illustrates both the demand for on-chain derivatives and the distortion introduced by subsidy-driven growth.
Centralized exchange volumes have moved in the opposite direction. The top 11 perpetual CEXs averaged $4.7 trillion in monthly volume during 2026, down from $7.1 trillion in 2025, according to TokenInsight's Q2 2026 exchange report. The contraction reflects macroeconomic headwinds — notably U.S. Treasury yields above 5% — rather than competitive displacement by DEXs.
Five platforms account for the majority of on-chain perpetual volume. Their architectures differ materially.
Hyperliquid operates a purpose-built Layer-1 blockchain with HyperBFT consensus, delivering sub-second finality (~0.2 seconds) and throughput of up to 200,000 orders per second. HyperCore runs a fully on-chain central limit order book (CLOB). HyperEVM, launched on mainnet in February 2025, added general-purpose smart contract programmability, enabling third-party applications like PaperTrade and OpenETF to build on top of the exchange infrastructure.
Open interest: $8.3–9.2 billion. Weekly volume (as of October 3, 2026): $43.8 billion. Cumulative revenue: $1.4 billion. Q3 2026 gross protocol revenue: $216.4 million, with $170.3 million from perp fees. The HYPE token traded near $86–93 in early October 2026, just below its all-time high of $98 reached September 23.
Aster emerged from the March 2025 merger of Astherus and APX Finance. It deploys across BNB Chain, Ethereum, Solana, and Arbitrum, offering dual-mode trading: a simplified interface with up to 1,001x leverage and a professional-grade CLOB with MEV protection and hidden orders.
In March 2026, Aster launched its own Layer-1 network. The platform has processed over $650 billion in cumulative volume, accumulated $450 million in TVL, and attracted 350,000 active wallets. Its 30-day perp volume in September 2026 was approximately $40 billion with $1.8 billion in open interest.
The volume-to-TVL ratio reached 70:1 during the September 2025 surge. A healthy range for derivatives platforms is typically 3–7:1, per 21Shares analysis, raising questions about organic demand.
Lighter is a custom zero-knowledge rollup built on Ethereum, founded by former Citadel Securities engineers. It runs an application-specific zk-rollup for order book trading with zero trading fees for retail users and sub-second finality.
The platform operates across four chains: Arbitrum, Ethereum, its native zkLighter, and Robinhood Chain. As of mid-2026, 30-day volume sat at roughly $35.7 billion, with $1.53 billion in open interest and $1.4 billion in TVL. The zero-fee model subsidizes growth through the expectation of future token economics and protocol-level monetization.
EdgeX has processed $515 billion in cumulative volume across 191,000 active addresses. Daily volume averaged $6.6 billion with monthly throughput of approximately $91 billion. The platform handles up to 200,000 orders per second. It runs on StarkWare infrastructure, which collectively powers approximately 16% of total on-chain perpetual volume.
A newer entrant, Variational reported $2.5 billion in cumulative volume, $33 million TVL, and $195 million in open interest as of early 2026. It recorded $19.5 billion in seven-day volume in the week ending October 3, 2026 — a notable surge that positioned it second behind Hyperliquid for that period. The protocol has raised $11.8 million in venture funding.
Fee structures vary significantly, creating distinct economic profiles.
| Platform | Maker Fee | Taker Fee | Revenue Model | |---|---|---|---| | Hyperliquid | 0.015% | 0.045% | 95% of fees fund HYPE buybacks | | Aster | 0.005% | 0.040% | 99% of daily fees buy ASTER for veASTER stakers | | EdgeX | 0.012% | 0.038% | Token incentives | | Lighter | 0% | 0% | Zero-fee; monetization TBD | | Variational | Varies | Varies | Early-stage; venture-funded |
Hyperliquid's fee model is the most economically proven. In Q3 2026, the platform generated $216.4 million in gross revenue. Approximately 95% of trading fees — collected through HyperCore, Builder Codes, and HIP-3 markets — are directed to open-market HYPE token buybacks. As of October 2026, cumulative buybacks exceeded $1.26 billion. The first AQAv2 (Automated Quarterly Auction) payout of approximately $14.5 million in USDC was processed on October 3, routed through Coinbase and Circle.
Lighter's zero-fee model is an explicit trade-off: sacrifice near-term revenue for market share. The founders are betting that order flow captured today can be monetized later through premium features, margin lending, or token-based value capture. It mirrors the playbook Robinhood Markets used in equities — subsidize the product, build the user base, monetize the flow.
Aster's 99% fee-to-staker distribution is the most aggressive redistribution in the perp DEX market. Whether it attracts sustainable liquidity or merely mercenary capital depends on whether the ASTER token maintains value independent of yield.
The concentration metrics tell a story of dominance with rising fragmentation.
By Open Interest (as of October 2026):
By 30-Day Volume (as of early October 2026):
Hyperliquid's share has compressed from approximately 70% in 2024 to 30.9% of weekly volume by October 2026. Its open interest share (56.8%) remains significantly higher than its volume share, indicating that while newer platforms capture speculative flow, Hyperliquid retains stickier capital with longer holding periods.
The 48% "others" category includes dYdX, GMX, Jupiter, Drift, Paradex (600+ markets, $1.47B daily volume, $796M open interest), and dozens of smaller venues. This long tail suggests the market is large enough to support multiple architectures, but no second platform has established the durable liquidity moat that Hyperliquid holds.
Volume data in perpetual DEXs requires careful interpretation. Several factors inflate reported figures relative to economically meaningful activity.
Aster's volume-to-TVL ratio reached 70:1 during its September 2025 surge, compared to a healthy range of 3–7:1, per 21Shares analysis. This does not prove wash trading, but it indicates that a significant portion of volume was generated by traders rotating capital at high leverage rather than deploying new risk.
The PaperTrade incident on October 10–11, 2026, demonstrated how thin order books on newer platforms create manipulation vectors. Two wallets used approximately $20 million in trades to shift Ethereum prices on Hyperliquid's BBO by 10–20 basis points, while maintaining nine-figure long positions on PaperTrade — which used Hyperliquid's midpoint price as its execution reference. PaperTrade tightened position limits in response.
Zero-fee platforms like Lighter face a structural challenge: without trading costs, the barrier to wash trading is zero. The platform counters that its zk-rollup architecture provides full auditability, but independent volume verification remains difficult.
CoinGecko's 2026 State of Crypto Perpetuals Report noted that Binance and OKX combined hold 48% of centralized perp market share, with DEX open interest share nearly quadrupling year-over-year. The report did not attempt to adjust DEX figures for incentivized or non-organic volume.
Centralized exchanges have not conceded the derivatives market passively.
Coinbase completed its $2.9 billion Deribit integration in October 2026, building a full derivatives stack that combines a Futures Commission Merchant, a Designated Contract Market (after CFTC granted a Derivatives Clearing Organization license), and international options infrastructure. The company plans to link U.S. and international liquidity pools.
Binance maintained 36.5% of centralized derivatives market share in Q1 2026, processing approximately $69 billion in daily perpetual volume. OKX held 16.4%, followed by Bybit (10.1%) and MEXC (9.5%). These four platforms controlled over 70% of centralized derivatives, per TokenInsight.
The CEX advantage remains latency (microsecond vs. millisecond execution), deeper liquidity, and broader margin asset support. The DEX advantage is self-custody, censorship resistance, and — in many cases — higher available leverage (up to 1,001x on Aster versus typical 100–125x on CEXs).
Morgan Stanley's addition of live Bitcoin, Ethereum, and Solana trading to E*TRADE signals that traditional brokerages now compete for the same flow. The derivatives market is fragmenting across three tiers: institutional (CME, Deribit/Coinbase), retail centralized (Binance, OKX), and on-chain (Hyperliquid, Aster, Lighter).
The on-chain perpetual futures market has grown from a rounding error to a $600+ billion monthly market in under three years. That trajectory, even adjusted for incentive-driven inflation, represents a structural shift in where derivatives execution occurs. Hyperliquid has the widest moat — measured by open interest concentration, revenue generation, and protocol maturity — but its volume share compression from 70% to 31% indicates that the moat is liquidity-deep rather than competition-proof. Five to seven platforms now compete for meaningful share, each with distinct architectural bets: custom L1 (Hyperliquid), multi-chain deployment (Aster), zk-rollup (Lighter), and StarkWare-based execution (EdgeX, Paradex).
The question is not whether on-chain perps are viable — that has been answered. The question is whether current volumes persist after token incentive programs mature, whether fee compression toward zero is sustainable, and whether the 12% market share ceiling is a plateau or a waypoint. Centralized venues retain structural advantages in latency and liquidity depth. The perp DEX segment has proven it can acquire traders. Retaining them without subsidies is the next test.