News Square
Global Crypto News · Real-time Aggregation
📚 Deep Analysis
More →NEAR enters crypto’s top 20 after 140% surge in 30 days
NEAR climbed into the top 20 cryptocurrencies by market capitalization, as NEAR Intents surpassed $31 billion in cumulative transaction volume. NEAR Protocol’s native token entered the top 20 cryptocurrencies by market capitalization on Thursday after rising $138% in the last 30 days, according to <a href="https://x.com/coingecko/status/2108022680628682828" rel="nofollow noopener" target="_blank">>CoinGecko</a>. At the time of writing, NEAR <a href="https://www.coingecko.com/en/coins/near" rel="nofollow noopener" target="_blank">>traded</a> around $5.51, up about 7% over 24 hours, with a market cap of about $7.2 billion and 24-hour trading volume reaching roughly $1.2 billion. NEAR ranks ahead of Stellar (XLM), which has a market cap of nearly $7 billion. NEAR Intents, a cross-chain trading protocol, has <a href="https://explorer.near-intents.org/" rel="nofollow noopener" target="_blank">>handled</a> over $31 billion in cumulative transaction volume, according to the NEAR Intents Explorer. NEAR’s rise in the rankings also <a href="https://cointelegraph.com/news/bitwise-launches-first-us-spot-near-etf-after-tokens-recent-surge" rel="nofollow" target="_blank">follows the Sept. 29 launch</a> of the first US spot exchange-traded product tracking the token. The Bitwise fund trades on NYSE Arca under the ticker NRR, offering investors exposure to NEAR without directly holding the crypto. NEAR Intents <a href="https://cointelegraph.com/news/near-intents-recovers-entire-stolen-38m-after-ultimatum-to-exploiter" rel="nofollow" target="_blank">suffered a $3.8 million exploit</a> involving its deposit and withdrawal infrastructure last week. On Oct. 2, General Manager Alex Shevchenko said that all stolen funds had been returned, a day after the breach was disclosed. <em><strong>Related: </strong></em><a href="https://cointelegraph.com/markets/near-price-surge-intents-zcash-privacy" rel="nofollow" target="_blank"><em><strong>NEAR jumps nearly 80% in a week as Intents volume nears $30B</strong></em></a>
Circle brings USDC, EURC payments to SAP customers through Tereina
The partnership with SAP-backed Tereina will let businesses send and receive Circle’s stablecoins from within the financial software they already use. Circle is bringing its USDC and EURC stablecoins into enterprise payment workflows through a partnership with Tereina, a financial services company backed by German enterprise software giant SAP. The <a href="https://www.circle.com/pressroom/tereina-an-sap-backed-company-and-circle-bring-usdc-and-eurc-into-enterprise-workflows-starting-with-the-sap-ecosystem-behind-84-of-global-commerce" rel="nofollow noopener" target="_blank">>integration</a> will allow businesses using SAP’s financial software to send and receive USDC and EURC within the applications they already use to manage payments. USDC will be the preferred stablecoin for dollar-denominated transactions, while EURC will be available for euro-denominated payments. <em>Source: </em><a href="https://x.com/circle/status/2107803627477709148" rel="nofollow noopener" target="_blank">><em>Circle</em></a> Tereina provides payment infrastructure that companies can embed into business software, allowing them to make payments without switching to a separate financial platform. Tereina and Circle said they plan to test the stablecoin integration with customers over the coming months, including for global payments and treasury operations. SAP’s ecosystem generates 84% of global commerce, according to the announcement, giving the integration a potentially broad enterprise reach. The move comes less than a month after <a href="https://cointelegraph.com/news/circle-arc-mainnet-launch-usdc-native-gas" rel="nofollow" target="_blank">Circle launched the mainnet of Arc</a>, its layer-1 blockchain focused on stablecoin payments and financial markets. Arc uses USDC for transaction fees and supports more than 20 fiat-backed stablecoins, including USDC and EURC. <em><strong>Magazine: </strong></em><a href="https://cointelegraph.com/magazine/too-big-to-pause-the-economic-stakes-of-an-ai-slowdown" rel="nofollow" target="_blank"><em><strong>Too big to pause: Could an AI slowdown crash the economy?</strong></em></a>
Crypto Late News | CFTC Chair: Failure of Clarity Act Prompts New Crypto Rules
Crypto News October 8: 1. CFTC Chair: Failure of the Clarity Act prompts new crypto rules; 2. U.S. SEC approves 3x leveraged ETFs involving Bitcoin and Ethereum; 3. CFTC proposes first round of crypto market regulatory rules, clarifying the framework for BTC and ETH; 4. CFTC proposes new rules to regulate the U.S. crypto market; 5. CFTC proposes a crypto market structure framework, exploring a voluntary federal registration path; 6. German regulator rejects Futurum Bank's MICA application, bitcoin.de plans a new model; 7. Germany's Federal Financial Supervisory Authority rejects futurum bank ag's MICA license application, bitcoin.de trading suspended; 8. Hong Kong plans to submit a bill to regulate crypto brokers and custodians; 9. German regulator rejects Bitcoin Group SE's MiCA license application; 10. CFTC proposes crypto rules, does not cover spot trading.
Perplexity open-sources a new retrieval model: a 0.6B small model can directly query an index built by a 9B model.
Perplexity has open-sourced two multimodal embedding models, pplx-embed-v2-late, at 0.6B and 9B respectively. The new models support text, image, and PDF page retrieval, and the vectors of the two models are mutually compatible, allowing developers to build a knowledge base with the 9B large model and then use the 0.6B small model for everyday searches. The new models retain a 128-dimensional vector for each token, enabling better matching of relevant content within documents. In the official Vidore V3 image retrieval test, it was found that using the 0.6B model for both database construction and querying yielded a score of 62.3%; switching to the 9B model for database construction and the 0.6B model for querying raised the score to 63.5%. Using the 9B model entirely yielded a score of 65.2%. The weights of both models have been released on Hugging Face under the MIT license.
Kyle Rodda: Bitcoin rebound reignites debasement trade
According to CoinWorld news, Capital.com analyst Kyle Rodda said that Bitcoin's rebound is heading toward the $86,000 target, reflecting a recovery in safe-haven demand against fiat currencies. Rodda linked the rally to Scott Bessent's intervention in the Treasury market, arguing that Bitcoin is increasingly being used as a hedge against expansive U.S. policy and as a portfolio diversification option. He said the market outlook is improving and the cyclical low may already be in the past.
The U.S. Treasury Department withdraws two controversial cryptocurrency surveillance proposals.
Coin World News: The U.S. Treasury Department has withdrawn two long-pending Financial Crimes Enforcement Network (FinCEN) proposals targeting unhosted crypto wallets and cryptocurrency mixing services. The wallet proposal would have imposed reporting and recordkeeping requirements on certain transactions exceeding $3,000 and $10,000, while the mixing proposal sought to expand reporting on transactions that may involve foreign crypto mixers. The withdrawal formally ends the rulemaking process for both proposals.
AMD off-exchange short sale volume ratio reported at 64.12%
Bjie.com US stock data: AMD's off-exchange short selling volume accounted for 64.12% that day. Compared with the same-caliber snapshot 24 hours earlier, it rose by 3.8766 percentage points.
The protocol's fees over the past 24 hours were reported at $75.2713 million.
CoinWorld data: Within the period, the protocol's fees over the past 24 hours were 75.2713 million USD, an increase of 1.28% compared to the previous statistical period under the same methodology.
US options market volume was 13,632,317 contracts.
CoinWorld US stock data: US options market volume was 13,632,317 contracts, of which 7,309,174 were calls and 6,323,143 were puts.
25 models collectively lost to humans: AI still cannot understand many everyday common sense concepts.
According to news from Bijie.com, as reported by OneMillion_AI, a new evaluation released by Scale AI's research institution Scale Labs in collaboration with Elorian shows that 25 multimodal models collectively lost to humans in testing, failing to understand many common-sense aspects of daily life. The study covered 522 open-ended questions, involving 288 images and 234 videos, and human participants achieved an accuracy rate of 93.1%. The top-ranked GPT-6 Astra achieved only 53.6% accuracy, while GPT-6.1 SOL and Claude Opus scored 46.6% and 44.6%, respectively. The study found that 94% of model failures were related to missing key clues, misidentifying objects, or being unable to infer implicit relationships. In terms of social understanding, 21 models performed the worst, and video questions were generally more difficult than image questions. Despite increased reasoning, model performance still failed to catch up with humans.