Solana’s PumpFun Hits $3M in Daily Revenue, Flips Hyperliquid

TheCryptoTimes发布于2025-09-15更新于2025-09-15

The Solana-based memecoin platform PumpFun is once again gaining traction, hitting $3 million in daily revenue for the first time since February. With this, it has flipped one of the most popular crypto perpetuals trading platform Hyperliquid. 

The surge in PumpFun’s revenue comes amid renewed interest in memecoin trading. Besides, PumpFun’s recently launched token buyback program is also driving huge activity back to the platform. 

As per DeFillama data, PumpFun generated $3.12 million in the past 24 hours, making it the third largest revenue generating crypto platform after Tether and Circle. It has surpassed Hyperliquid, which has constantly been making headlines since the past few months. 

Top Crypto Protocols Ranked By Revenue ChartTop Crypto Protocols Ranked By Revenue Chart
Source: DeFiLlama

Third-largest protocol by revenue generation

PumpFun’s daily revenue peaked at $7.07 million on January 23, when memecoins were at the center of attention within the crypto industry. The platform has so far generated $807.8 million in cumulative revenue since its launch in January 2024. 

Positioned third after stablecoin issuers Tether (USDT) and Circle (USDC), generating over $21.7 million and $7.62 million in revenue, respectively. PumpFun has been constantly dominating this list among hundreds of crypto platforms. Its most of the revenue comes from the fees paid by users for trading and launching tokens fees. 

Also Read: Shibarium Bridge Loses $2.4M in Flash Loan Exploit


Mobile Only ImageMobile Only Image

热门币种推荐

你可能也喜欢

拖更三年,北大校友翁荔最新长文刷屏

前OpenAI副总裁翁荔(Lilian Weng)发表了一篇关于AI扩展定律(Scaling Laws)的深度分析文章。文章指出,这条指导了AI行业数百亿美元投入的核心定律,远比人们想象的更为脆弱。 文章回顾了Scaling Laws的基本思想,即模型性能随规模扩大而可预测地提升。然而,OpenAI与DeepMind在关键问题上得出了相反结论:给定算力,资源应更多分配给模型还是数据?OpenAI的Kaplan团队认为模型增长应更快,而DeepMind的Chinchilla团队则认为应等比增长。后来研究发现,这一分歧源于参数统计口径的差异和实验规模不足,导致Kaplan的结论仅适用于小规模场景。 更关键的是,被行业广泛采纳的Chinchilla最优配比公式本身也存在方法论瑕疵。2024年有团队复现发现,其损失函数因取均值而非求和,导致优化器提前停止,输出并非最优解。此外,用于外推的关键参数精度不足,放大了误差。 文章进一步指出,经典Scaling Laws的根本前提——高质量数据无限供应——正在崩塌。数据重复训练不可避免,新研究引入了惩罚项来修正公式,并发现大模型对数据重复更敏感。这解释了行业为何转向强化学习、测试时计算和合成数据等新路径。 翁荔的博客以其清晰深入的技术解析著称,这篇文章历时三年完成。她于2025年联合创立了新公司Thinking Machines Lab。文章强调,下一代AI的进步不仅依赖算力规模,更取决于对这些基础定律细节更精确的理解与运用。

marsbit54分钟前

拖更三年,北大校友翁荔最新长文刷屏

marsbit54分钟前

交易

现货
合约

热门文章

相关讨论

欢迎来到HTX社区。在这里,您可以了解最新的平台发展动态并获得专业的市场意见。以下是用户对SOL(SOL)币价的意见。

活动图片