Perpetual DEX Volume Hit Record High $898 Billion in Q2 2025: CoinGecko

TheCryptoTimesPublicado em 2025-07-17Última atualização em 2025-07-17

In the second quarter of 2025, perpetual trading on decentralized exchanges (DEX) soared to an $898 billion in volume, according to CoinGecko. Hyperliquid was at the forefront of this trend, taking 72.7% of the market share. It recorded $653.2 billion in trading volume and ranked eighth among all perpetual trading platforms, including centralized exchanges.

The report states that decentralized exchanges saw record growth in both spot and perp volumes. Most of it was driven during the Hyperliquid wave, where the exchange saw significant traction.

Besides Hyperliquid, only three other perp DEXs saw volume growth: Aster, RabbitX, and EdgeX. Aster, formerly APX Finance, saw volume double after launching its Pro mode. Meanwhile, dYdX continued to slide. Its average monthly volume dropped to $5.3 billion—half of January’s figure.

Crypto Market Recovers, But Centralized Volumes Fall

The crypto market made a comeback in Q2, with total market capitalization soaring by 24.0% to reach $3.5 trillion. Bitcoin was at the forefront of this resurgence, breaking through $100,000 and hitting a new all-time high. Its market dominance climbed to 62.1%, drawing in the majority of the capital inflow. However, spot volumes on centralized exchanges fell again, down 27.7% QoQ to $3.9 trillion.

In contrast, spot volumes on decentralized exchanges rose 25.3% QoQ. This spike pushed the DEX:CEX volume ratio to an all-time high. PancakeSwap led the spot DEX sector, benefiting from the wider shift.

Also Read: Solana DEX Raydium’s RAY Token Buybacks Cross $190 Million

Follow The Crypto Times on Google News to Stay Updated!

Leituras Relacionadas

How Ethereum’s ‘Lean’ roadmap could reshape ETH over the next 4 years

Ethereum has entered a new development phase with the introduction of the 'Lean Ethereum' roadmap, a multi-year overhaul aiming to reshape the network's core protocol over the next 3-4 years. The plan focuses on major technical upgrades to improve scalability, security, and long-term resilience. Key components include implementing recursive STARKs, quantum-safe cryptography to guard against future threats, a multidimensional gas model, and a redesigned state architecture using Verkle trees. These changes aim to significantly increase transaction throughput, lower costs, and improve finality speed while maintaining backward compatibility. Infrastructure development is already underway, with initiatives like post-quantum secure devnets and the Glamsterdam upgrade laying the groundwork. Efforts are also focused on scalable state management and increasing the network's gas limits. The overarching goal is to future-proof Ethereum, enabling it to scale efficiently and securely throughout the next decade with minimal disruption to existing applications.

ambcryptoHá 1h

How Ethereum’s ‘Lean’ roadmap could reshape ETH over the next 4 years

ambcryptoHá 1h

Germany’s banking giants bring crypto to 80M customers – Why it matters

Germany's two largest banking networks, Sparkassen and Volksbanken Raiffeisenbanken, are integrating cryptocurrency trading directly into their banking apps, potentially reaching around 80 million customers. This marks a major shift as these groups previously deemed crypto too risky. They are building their own regulated infrastructure, such as the BaFin-approved "meinKrypto" platform, rather than partnering with external exchanges. While this integration brings crypto to a vast retail audience within a trusted framework, critics warn it could lead inexperienced investors to underestimate the high volatility and risks. Despite a recent decline in retail crypto trading volume, Germany remains a top-ten market. The move signifies a significant step in merging traditional banking with digital assets in Europe.

ambcryptoHá 2h

Germany’s banking giants bring crypto to 80M customers – Why it matters

ambcryptoHá 2h

Aerodrome Finance: Can AERO target $0.65 after a 22% weekly rally?

Aerodrome Finance (AERO) has rallied strongly, gaining 22% in the past week. After holding support at $0.45, the altcoin reached $0.60 for the first time since January and was trading at $0.58 at the time of writing. Technical indicators show bullish momentum, with the RSI at 67 and the DMI Modified indicator rising sharply, suggesting a strengthening uptrend. This price movement is supported by significant market demand. Open Interest in derivatives surged 10% to $56 million in a day, and spot trading has shown a positive buy-sell delta over recent days, with net exchange flow turning negative, indicating accumulation. Analysts suggest that if bullish pressure continues, AERO could target $0.65. However, a failure to sustain above $0.60 could weaken the trend, potentially leading to a retest of the $0.46-$0.50 support zone.

ambcryptoHá 3h

Aerodrome Finance: Can AERO target $0.65 after a 22% weekly rally?

ambcryptoHá 3h

Why the UXLINK hacker’s 14,336 ETH transfers raise fresh questions for DeFi

Recent on-chain activity reveals the UXLINK exploiter is actively laundering stolen funds, complicating tracing efforts. The September 2025 hack, which exploited a 'delegateCall' vulnerability to steal roughly $4.5 million, involved converting illicit assets into DAI and Ethereum. Recently, the attacker swapped over 10 million DAI for approximately 6,000.8 ETH and has deposited a total of 14,336.6 ETH into the privacy mixer Tornado Cash in recent weeks, including over $8.1 million worth in a single move. Simultaneously, wallets connected to the defunct Mining Express Ponzi scheme have been converting long-held assets, swapping 5,004 ETH for 8.8 million DAI. Following a separate exploit of the Jaredfromsubway.eth MEV bot, about $5.1 million was also routed to Tornado Cash. These incidents underscore a critical vulnerability in DeFi: while enabling permissionless transfers, the ecosystem lacks effective mechanisms to halt or manage illicit funds once in motion, allowing them to be moved and concealed relatively easily. This highlights the need for strengthened cross-network coordination and real-time threat detection to better safeguard the space.

ambcryptoHá 5h

Why the UXLINK hacker’s 14,336 ETH transfers raise fresh questions for DeFi

ambcryptoHá 5h

Li Fei-Fei's Latest Long-Form Article: When Video Generation, Robotics, and NVIDIA All Call Themselves World Models, We Need a Taxonomy

In a new article, Dr. Fei-Fei Li addresses the widespread and often inconsistent use of the term "world model" in AI. She proposes a clear, functional taxonomy rooted in the classic Partially Observable Markov Decision Process (POMDP) loop (agent → action → state → observation → agent). According to this framework, current systems called "world models" are different projections of this loop, categorized by their primary output: 1. **Renderers**: Output observations (pixels). Their goal is visual fidelity for human consumption (e.g., video generation models like Sora). They are the most commercially mature but are limited by a focus on appearance over physical accuracy. 2. **Simulators**: Output states (geometric, physical, dynamic representations). They provide a structurally accurate world for both human professionals (e.g., architects) and computational agents (e.g., robots for training). Li argues simulators are the crucial, underappreciated bridge, as they can underpin both rendering and planning. 3. **Planners**: Output actions. Given an observation and a goal, they decide what an agent should do next (e.g., robotic action models). This area is highly promising but remains the least mature for real-world deployment. Li highlights a key trend: the boundaries between these three categories are beginning to blur, as they all rely on a shared underlying understanding of geometry, physics, and dynamics. The logical endpoint is a unified world foundation model capable of switching between rendering, simulation, and planning based on downstream needs. This convergence, she concludes, is central to advancing spatial intelligence—enabling machines not just to talk about the world, but to truly understand, imagine, and interact with it.

marsbitHá 6h

Li Fei-Fei's Latest Long-Form Article: When Video Generation, Robotics, and NVIDIA All Call Themselves World Models, We Need a Taxonomy